VIScorp's possible plans of using the DEC Alpha in some form
George Noel
In article <4rdlkc$2...@info.htcomp.net>,
Tony Belding <tlbe...@htcomp.net> wrote:
> RD> From: Ross Deeley <ros...@ramhb.co.nz>
>
> RD> But that would require the programs to be recompiled for each CPU.
>
>I consider this a huge problem. To see the Amiga OS splintered that way would be extremely Bad News for everyone involved. I hope VIScorp understand this.
Funny, I don't see Microsoft complaining with Windows NT being on so many
processors. If VIScorp are seriously considering this, why not wait and
hear their plans and aims for the future before condemming it.. I doubt
if AmigaOS was ported to the DEC Alpha that it or the machine running it
would be aimed at the lowend/home user. The ED/68K Amiga and PowerPC
Amiga compatibles would suit there nicely. The DEC Alpha Amigas could be
aimed at the highend for now unless DEC comes up with a stripped down
version of their processors/cost reduce them to be used in a home market
PC environment. Bill Buck said they were considering it to "improve the
video" anyways, he didn't mention it in regards to being the main CPU.
Maybe they have it in mind to create a graphical chipset with a DEC Alpha
core like Commodore was doing with Hombre having a HP core.. like
Nintendo have done using an SGI core. Normally it would be a CPU from the
same company to go along with the graphical chips etc. but heck, look at
the Atari Jaguar.. it has two 64 bit RISC graphic and blitter chips which
are the main two chips of the system but the CPU? A 68000, 16 BIT CPU
(mainly there to tell the other processors what to do and to read the
joystick ports :) ) so why couldn't VIScorp have a DEC Alpha core graphic
chipset (Carl Sassenrath did say they were talking to R.J. Mical [one of
the orginal Amiga chipset designers] and that some ideas were tossed
around) and still use a PPC CPU?
Carl wants to build a "Killer Amiga", will we get a "Killer Amiga" using
the same chips as Apple? No... mind you, I am sure the Amiga version
would be better anyways and better in price/performance then the Intel
line due to the new choice of hardware as standard, increased Mhz for the
CPU and the AmigaOS.
Right now a majority of home computers are running on 32 bit hardware, so
are the game consoles (Atari [debateable] and Nintendo [soon] aside).
The Amiga should look forward to the future and be the Nintendo of the
computer industry... putting any "games machine" put downs aside. The
best the lowend PPC Amiga compatibles will be is still 32 bit, like the
current Pentium. There might be a possibility the highend PPCP
compliant Amiga compatible would use the 620, the current ONLY CPU in the
PPC line that is 64 bit! Intel will be teaming with Hewlett Packard (HP)
in 1997 to develop their next generation CPU which will be a RISC design
though I am not sure if it will still be 32 bit or 64 bit. You also now
see at least 1 PC graphic chip/accelarator makers touting 128 bits. Why
not try to stay ahead by porting AmigaOS to a 64 bit processor and/or
using a 64 bit graphics sub-system as standard to stay ahead of the game?
If they can make this lowend, then perfect, they are keeping with the
Amiga philosphy, if not, make it highend, it will still be an Amiga and
would probably bring in new developers and create new niches. The old
developers can stick with what they are used to and go to the 32 bit PPC
path with hopes of the ONE (1) so far 64 bit CPU and have a choice to
also support a 64 bit total Amiga while the new developers develop for
the 64 bit total Amiga and if they chose, also develop for the PPC Amiga
compatibles once they see how powerful AmigaOS can be.
If they could get this 64 bit technology (don't forget DEC Alpha's START
at 64 bit) at a low cost, then imagine the set-top boxes they could make
then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
sure VIScorp could do amazing things with that kind of technology in a
set-top box and a computer!
> || Tony L. Belding, Hamilton Texas
-=*George*=-
Kolbjørn Barmen
I have always wondered about this..
From one of motorlas webservers, http://129.38.233.1/prod/0X0/68000.html :
MC68000 & Derivatives
(Including EC, LC, and HC)
NOTE: Motorola no longer makes the MC68000 only it's derivatives.
* 32-Bit Data and Address Registers
* 16-Mbyte to 6-Gbyte Direct
* Addressing Range
* Program Counter
* 56 Powerful Instructions
* Operations on Five Main Data Types
* Memory-Mapped Input/Output (I/O)
* 14 Addressing Modes
What's the 16 bit in here?
According to Motorola them selves, all the 68k family is 32-Bit.
Btw.. motorolas web-site is *excelent* !!
I can imagine intel give 5 years warranty on their products... NOT!
--
Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
Erik Trulsson
In comp.sys.amiga.misc Kolbjørn Barmen <ko...@nvg.unit.no> wrote:
> George Noel (gn...@morgan.ucs.mun.ca) wrote:
> > A 68000, 16 BIT CPU
>
> I have always wondered about this..
>
> From one of motorlas webservers, http://129.38.233.1/prod/0X0/68000.html :
>
>
> MC68000 & Derivatives
> (Including EC, LC, and HC)
>
> NOTE: Motorola no longer makes the MC68000 only it's derivatives.
>
> * 32-Bit Data and Address Registers
> * 16-Mbyte to 6-Gbyte Direct
> * Addressing Range
> * Program Counter
> * 56 Powerful Instructions
> * Operations on Five Main Data Types
> * Memory-Mapped Input/Output (I/O)
> * 14 Addressing Modes
>
> What's the 16 bit in here?
>
> According to Motorola them selves, all the 68k family is 32-Bit.
>
The data bus on 68000 is 16-bit. One very common question is: Is the
68000 32-bit or 16-bit. The answer to that depends. From a hardware point
of view the 68000 is a 16-bit CPU with a 24-bit address bus while from
a software point of view the 68000 is a 32-bit CPU with 32-bit addresses.
Kolbjørn Barmen
Erik Trulsson (trul...@Manama.docs.uu.se) wrote:
> The data bus on 68000 is 16-bit. One very common question is: Is the
> 68000 32-bit or 16-bit. The answer to that depends. From a hardware point
> of view the 68000 is a 16-bit CPU with a 24-bit address bus while from
> a software point of view the 68000 is a 32-bit CPU with 32-bit addresses.
Thanx for clearing this up!
-- kolla
Yoda
George Noel wrote:
>
>>I consider this a huge problem. To see the Amiga OS splintered that way
>>would be extremely Bad News for everyone involved. I hope VIScorp understand this.
>
> Funny, I don't see Microsoft complaining with Windows NT being on so many
> processors.
NT was designed to run on a range of CPUs - AmigaOS wasn't. To give the
Amiga equivalent functionality would require, at the very least, a complete
re-write of Exec. Before I proceed, let me state that I was at the Toulouse
meeting. I heard *nothing* that suggested to me that Viscorp was even remotely
interested in undertaking such a task themselves - all they said that
was they would "support" such endeavours (i.e. anyone who wants to
put AmigaOS on CPU X is welcome to do so, so long as Viscorp gets the
licensing/royalty fees). The only extensions and improvements to the AmigaOS
that they formally committed themselves to were those that would
make ED a better set-top box, not those that would make the Amiga
a better computer. Hell, at the Developers' closed afternoon session,
someone (I think he was from Clianto, the Personal Paint people)
requested them to state very clearly and formally whether they would or
would not choose the PowerPC as the next Amiga CPU, and whether
they would port AmigaOS to it. And all we got back was this parrot-
like assertion that they would "support" anyone who wanted to
port the Amiga OS to any CPU ... UGH!
>If VIScorp are seriously considering this, why not wait and
> hear their plans and aims for the future before condemming it.. I doubt
> if AmigaOS was ported to the DEC Alpha that it or the machine running it
> would be aimed at the lowend/home user.
I haven't been following this thread, so I wouldn't know where this
DEC Alpha AmigaOS port rumour comes from. At Toulouse, though, Viscorp
made a certain amount of fuss about how they were talking to "some big
companies", and DEC was explicitly cited as one of those big companies.
However, this was in connection with their stated aim of seeking some wealthy
partners for their set-top boxes. We were told, for instance, that "when
it comes to video [and, by extrapolation, to set-top boxes], Amiga's the
only game in town", and that is why DEC had presumably shown any interest
at all. However, with DEC now one of the official licensees of Oracle's NC
(Network Computer) design, it is debatable whether they would partner with
a startup like Viscorp.
> Bill Buck said they were considering it to "improve the
> video" anyways, he didn't mention it in regards to being the main CPU.
> Maybe they have it in mind to create a graphical chipset with a DEC Alpha
> core like Commodore was doing with Hombre having a HP core.. like
> Nintendo have done using an SGI core.
Not a chance in hell! During the public Q&A session, someone asked if they
had any plans to resurrect Hombre. The reply was that Ed Hepler had been
in touch with them, but that there was no need to invest in custom design
when there were off-the-shelf parts like the Phillips Trimedia chip. Carl
Sassenrath himself stated, and I quote, "those days are over", a sentiment
that was later repeated by David Rosen.
(OK - before the anti-custom chip brigade gets too excited about this,
let me point out that the idea of incorporating special 3D instructions
into a RISC core is where the mainstream is heading, with Intel's MMX
technology and the HP/Intel alliance. Of course, Commodore engineers
*almost* got there first. Without this kind of vision and foresight, that
only true Amigans like Jay Miner and Ed Hepler could have injected into
their work, there is little need for a non-Wintel computer like the Amiga).
> Carl wants to build a "Killer Amiga"
Yes - Bill Buck repeated this line at the close of the meeting, and it
was included in the IRC chat that was going on simultaneously. What that
IRC script did not include were the words Bill Buck used afterwards. It
went something like this:
"In the words of Carl Sassenrath, I want to build a killer
Amiga. The ED is one killer Amiga"
There is only one way of interpreting that last sentence, and it is
not a very pro-Amiga one :-(.
I don't wish to dampen people's enthusiasm about Viscorp, but it is
important that hopes are not unnecessarily raised. The only positive
thing that I have seen since Toulouse has been the statement from
Viscorp that they wanted to be a player in the Amiga *computer*
market. Guess we'll just have to wait and see. And if Viscorp turns
out to be another Escom-like predator who wants its greedy hands on
our beloved Amy, then there's always Phase 5 :-)
> -=*George*=-
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: If they could get this 64 bit technology (don't forget DEC Alpha's START
: at 64 bit) at a low cost, then imagine the set-top boxes they could make
: then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
: sure VIScorp could do amazing things with that kind of technology in a
: set-top box and a computer!
Personally, I would rather they went with parrallel cpus before using 64 bit
cpus. Then again it depends on what you call 64 bit too.
Regards,
BM
Lynn Winebarger
I have heard from a local dealer that DEC will take a binary and convert it
to an alpha binary free of charge for developers who want to change CPUs.
Of course, I don't know what the truth value of this information is, but
the dealer does sell DEC Alpha systems (for rendering) and has been fairly
reliable in the past.
If this were true, it would make the software side of the conversion
quite a bit easier.
Lynn
John S. Burger
On or about 04-Jul-96 04:43:37 Kolbjørn Barmen typed the following words
about "Motorola and the m68k (was "Re: VIScorp's possible plans..")". My reply is thus...
Hi Kolbjørn,
KB> George Noel (gn...@morgan.ucs.mun.ca) wrote:
>> A 68000, 16 BIT CPU
KB> I have always wondered about this..
KB> From one of motorlas webservers, http://129.38.233.1/prod/0X0/68000.html
KB> :
KB> MC68000 & Derivatives
KB> (Including EC, LC, and HC)
KB> NOTE: Motorola no longer makes the MC68000 only it's derivatives.
KB>
KB> * 32-Bit Data and Address Registers
KB> * 16-Mbyte to 6-Gbyte Direct
KB> * Addressing Range
KB> * Program Counter
KB> * 56 Powerful Instructions
KB> * Operations on Five Main Data Types
KB> * Memory-Mapped Input/Output (I/O)
KB> * 14 Addressing Modes
KB> What's the 16 bit in here?
Where does the above say anything about 16 bit???
KB> According to Motorola them selves, all the 68k family is 32-Bit.
KB> Btw.. motorolas web-site is *excelent* !!
KB> I can imagine intel give 5 years warranty on their products... NOT!
KB> --
KB> Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
KB> ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
--
-= John =- jsbu...@xmission.com BIX: jburger
Amiga 2000 Via
68030 25MZ Amateur Radio KB0ES T
2+44 MB RAM // H
850 MB HD John S. Burger Hooper, UT \X/ O
Iomega ZIP R
Toshiba 6.7x CD This message was composed on... 2.31
4-Jul-96 10:38:45 MDT
--
Used Iraqi rifles for sale: Dropped once, never fired...
Brian Quinlan
trul...@Manama.docs.uu.se (Erik Trulsson) writes:
>The data bus on 68000 is 16-bit. One very common question is: Is the
>68000 32-bit or 16-bit. The answer to that depends. From a hardware point
>of view the 68000 is a 16-bit CPU with a 24-bit address bus while from
>a software point of view the 68000 is a 32-bit CPU with 32-bit addresses.
The 68000 can be characterised as a 16-bit implementation of a 32-bit
processor series.
--
Brian Quinlan "Never ask what sort of computer a guy drives. If he's a Mac
qui...@sfu.ca user, he'll tell you. If not, why embarrass him?" - Tom Clancy
Donald Dalley
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: compliant Amiga compatible would use the 620, the current ONLY CPU in the
: PPC line that is 64 bit! Intel will be teaming with Hewlett Packard (HP)
: not try to stay ahead by porting AmigaOS to a 64 bit processor and/or
: using a 64 bit graphics sub-system as standard to stay ahead of the game?
: If they can make this lowend, then perfect, they are keeping with the
: Amiga philosphy, if not, make it highend, it will still be an Amiga and
: would probably bring in new developers and create new niches. The old
: developers can stick with what they are used to and go to the 32 bit PPC
: path with hopes of the ONE (1) so far 64 bit CPU and have a choice to
: also support a 64 bit total Amiga while the new developers develop for
: the 64 bit total Amiga and if they chose, also develop for the PPC Amiga
: compatibles once they see how powerful AmigaOS can be.
: If they could get this 64 bit technology (don't forget DEC Alpha's START
: at 64 bit) at a low cost, then imagine the set-top boxes they could make
: then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
: sure VIScorp could do amazing things with that kind of technology in a
: set-top box and a computer!
: -=*George*=-
Very well put, George! I filled in the VIScorp questionaire, with fewer
words, saying that they should get the OS into 64-bit mode. If 64 bit
chips/OS were used, the Amiga would definately have an edge in timing and
technology. This may be just the ticket.
I'd like to see some discussion re: a complete 64-bit Amiga.
Donald
--
+ Are you experienced? Observant?
+ Send your subway/bus rider tips to ab...@torfree.net .
Michael Kraemer
In article <4rgap9$d...@due.unit.no>, ko...@nvg.unit.no (Kolbjørn Barmen) writes:
> George Noel (gn...@morgan.ucs.mun.ca) wrote:
> > A 68000, 16 BIT CPU
>
>
>
> I can imagine intel give 5 years warranty on their products... NOT!
>
They don't need to. At the rate they are pumping out new x86 generations
it's not worthwhile to get a replacement. If your '486 breaks, dump it
and get a '586 instead. Motorola in contrast needs/needed much more time to
develop their 68k CPUs. Just look how many years they needed for the 68040
and 68060. You still can buy 68020/30/40 chips and probably also some
derivative of the 68000. Don't know if intel still makes 8086/80186/80286/80386,
probably not.
George Noel
In article <4rgap9$d...@due.unit.no>, Kolbjørn Barmen <ko...@nvg.unit.no> wrote:
>George Noel (gn...@morgan.ucs.mun.ca) wrote:
>> A 68000, 16 BIT CPU
>
>I have always wondered about this..
>
>
>
> MC68000 & Derivatives
> (Including EC, LC, and HC)
>
>
>What's the 16 bit in here?
>
>According to Motorola them selves, all the 68k family is 32-Bit.
>--
> Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
> ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
AARGH!! Major blunder on my part, of course the 68000 is a 32 bit CPU
(hope I don't get flamed too much for that mistake :) ). For a second
there I got mixed up in it being used in 16 bit systems such as the A500
and instead said the CPU was 16 bit. Heck and me for so long saying to
the PC crowd that our applications have been 32 bit from day one since
1985 due to the 68000 and they are only recently getting 32 bit apps.
Forgive me. :)
-=*George*=-
David Meyer
George Noel (gn...@morgan.ucs.mun.ca) wrote:
:
: Right now a majority of home computers are running on 32 bit hardware,
Not exactly true. The CPUs on nearly all Personal Computers - for home or
other use - have been 32-bit (or better) for years. That does not qualify
as "32-bit hardware". Sitting to my right is one of the last '486 models.
From a respected name, a low-end option to the early (and expensive)
Pentium models. 32-bit, you say? ISA bus, I say. And it is not/was not an
abberation. a lot of low-end models use this old and trusted 16-bit
architecture. Some brag about having a faster slot for their video cards
(imagine! haveing a discrete processor and memory for your video and
another for your audio! -radical-) but go with ISA for the rest.
David Meyer
Lynn Winebarger (owin...@nickel.ucs.indiana.edu) wrote:
: I have heard from a local dealer that DEC will take a binary and convert it
: to an alpha binary free of charge for developers who want to change CPUs.
: Of course, I don't know what the truth value of this information is, but
DEC has (had?) a facility dedicated to this process back when the Alpha
was introduced. It was on-line before the Alpha was actually commercially
available (in any form). The facility is/was free for the users, although
the users supplied the core of the staff used in the conversion. So, it
wasn't -exactly- free. Just really, -really-, low cost. This program may
have been either modified or eliminated in the intervening years.
dmax...@freenet.vcu.edu
VIScorp has just worked out an arrangement with Emerson Radio
to market their set-top box. YOu can read about it on the Amiga
Technology web site (and maybe on VIScorp's own site by now).
>
> RE: VISCORP'S POSSIBLE PLANS OF USING THE DEC ALPHA IN SOME FORM
>
> Newsgroups:
> comp.sys.amiga.misc,comp.sys.amiga.advocacy,comp.sys.amiga.hard
> ware,comp.sys.amiga.graphics,comp.sys.amiga.applications,comp.s
> ys.amiga.multimedia
> Post a followup article to newsgroup(s)
> From: ab...@torfree.net (Donald Dalley)
> Send e-mail reply to: Donald Dalley
> Date: Fri, 5 Jul 1996 03:37:27 GMT
> Organization: Toronto Free-Net
--
%%%%%%%%%%% less is more %%%%%%%%%%
%%%%% Don Maxwell, Richmond, VA %%%%%
%%%%%%%%% beauty is truth %%%%%%%%%
George Noel
In article <31DBDC...@revenge.jedi.com>,
Yoda <war...@revenge.jedi.com> wrote:
>George Noel wrote:
>>
>>>I consider this a huge problem. To see the Amiga OS splintered that way
>>>would be extremely Bad News for everyone involved. I hope VIScorp understand this.
>>
>> Funny, I don't see Microsoft complaining with Windows NT being on so many
>> processors.
>
>NT was designed to run on a range of CPUs - AmigaOS wasn't. To give the
>Amiga equivalent functionality would require, at the very least, a complete
>re-write of Exec. Before I proceed, let me state that I was at the Toulouse
Is this not what PIOS/VIScorp are doing? Once the code is ported to the
PPC, it shouldn't be that hard afterwards to tweak the compilation for
other processors.
>meeting. I heard *nothing* that suggested to me that Viscorp was even remotely
>interested in undertaking such a task themselves - all they said that
>was they would "support" such endeavours (i.e. anyone who wants to
>put AmigaOS on CPU X is welcome to do so, so long as Viscorp gets the
>licensing/royalty fees). The only extensions and improvements to the AmigaOS
There is such thing as contracting out. Also VIScorp mentions on their
web page under the Licensing section that one form of licensing is the
CO-Development licence or something to that effect. I would assume this
would go for hardware and software aspects.. such as the development on
the OS with PIOS.
>that they formally committed themselves to were those that would
>make ED a better set-top box, not those that would make the Amiga
>a better computer. Hell, at the Developers' closed afternoon session,
True but their developments will still make the Amiga a better computer
if they release their improvements to the computer range also. It just
might not be as big of a step we would like for them to take themselves.
>someone (I think he was from Clianto, the Personal Paint people)
>requested them to state very clearly and formally whether they would or
>would not choose the PowerPC as the next Amiga CPU, and whether
>they would port AmigaOS to it. And all we got back was this parrot-
>like assertion that they would "support" anyone who wanted to
>port the Amiga OS to any CPU ... UGH!
Well obviously VIScorp are not set on what to do with the technology yet
themselves but by other people doing some stuff for them and them getting
the money, as long as it fits in with their long term plans, then why
not? I mean, like i said above, PIOS porting AmigaOS to the PPC with the
help of VIScorp could then in return use that code for porting it to
other CPUs.
>>If VIScorp are seriously considering this, why not wait and
>> hear their plans and aims for the future before condemming it.. I doubt
>> if AmigaOS was ported to the DEC Alpha that it or the machine running it
>> would be aimed at the lowend/home user.
>
>I haven't been following this thread, so I wouldn't know where this
>DEC Alpha AmigaOS port rumour comes from. At Toulouse, though, Viscorp
There was a quote from Bill Buck himself from a speach transcribed at
Toulouse in which he stated they were looking at porting AmigaOS to the
DEC Alpha and that "that will help improve the video". This was later
passed off as a joke "to show how hard of a task it is to port AmigaOS to
other processors". Since then, Jason Compton, director of Communications
to VIScorp regarding the Amiga has stated there is still some indecision
inside of VIScorp as to what CPU to use and that the two main choices are
the PPC and Dec Alpha and that a decision will be made soon after the
buyout and we will hear about it then. Bill Buck said we will hear more
about the DEC Alpha situation in the middle of 1997.
>made a certain amount of fuss about how they were talking to "some big
>companies", and DEC was explicitly cited as one of those big companies.
>However, this was in connection with their stated aim of seeking some wealthy
>partners for their set-top boxes. We were told, for instance, that "when
License to make or to support hardware/software for them?
>it comes to video [and, by extrapolation, to set-top boxes], Amiga's the
>only game in town", and that is why DEC had presumably shown any interest
This is true, which is why I don't understand them wanting to get away
from the Amiga specific chipset and go with PC/off-the-shelf chips which
either A) don't have broadcast video as standard in chipsets or if video is
included, it is not broadcast quality. Have you seen the video from a
MAC? Talk about grainy!
>at all. However, with DEC now one of the official licensees of Oracle's NC
>(Network Computer) design, it is debatable whether they would partner with
>a startup like Viscorp.
True but VIScorp might be a recent "start-up" company but they are using
technology which has been established for years.
>> Bill Buck said they were considering it to "improve the
>> video" anyways, he didn't mention it in regards to being the main CPU.
>> Maybe they have it in mind to create a graphical chipset with a DEC Alpha
>> core like Commodore was doing with Hombre having a HP core.. like
>> Nintendo have done using an SGI core.
>
>Not a chance in hell! During the public Q&A session, someone asked if they
>had any plans to resurrect Hombre. The reply was that Ed Hepler had been
>in touch with them, but that there was no need to invest in custom design
>when there were off-the-shelf parts like the Phillips Trimedia chip. Carl
>Sassenrath himself stated, and I quote, "those days are over", a sentiment
>that was later repeated by David Rosen.
If "those days are over" then why the hell did they aquire the Amiga
technology? Talk about hypocrites! "When it comes to video, the Amiga is
where it is at" indicating off-the-shelf chips lack something yet they
run to those chips as "those days are over". If they are, then so is
this generation of ED's. As the saying goes "Nothing will get done, if
you don't do it yourself". If they are unhappy about off-the-shelf chips
and think "Amiga is where it is at" then obviously to do things right,
the way they want it to, they have to build the chips themself. "Those
days" ARE NOT over.. PC chip makers are still doing specific things for
PCs in their chips and constantly adding onto what was already there, why
can't VIScorp do the same with the Amiga chips? They still have a lot of
options which help animation and video that off-the-shelf parts do not,
use them and expand on them.
>(OK - before the anti-custom chip brigade gets too excited about this,
>let me point out that the idea of incorporating special 3D instructions
>into a RISC core is where the mainstream is heading, with Intel's MMX
>technology and the HP/Intel alliance. Of course, Commodore engineers
>*almost* got there first. Without this kind of vision and foresight, that
>only true Amigans like Jay Miner and Ed Hepler could have injected into
>their work, there is little need for a non-Wintel computer like the Amiga).
Yes, they were true visionaries/geniuses and least to mention
innovators. Dave Haynie and now VIScorp sound like a bunch of wimps
unwilling to take "riscs" and are copping out and doing what everyone else
is doing. This isn't what the Amiga is about. Believe in your product,
set out to do what no one else is doing or in a way that advances what
everyone else is doing and then market it and reap in the rewards without
having to pay anything to anyone else. This is how Sony got their
Playstation and its custom chips and this is why it is the number one
selling games machine out there now - they have the better specs for the
price.
>> Carl wants to build a "Killer Amiga"
>
>Yes - Bill Buck repeated this line at the close of the meeting, and it
>was included in the IRC chat that was going on simultaneously. What that
>IRC script did not include were the words Bill Buck used afterwards. It
>went something like this:
> "In the words of Carl Sassenrath, I want to build a killer
> Amiga. The ED is one killer Amiga"
>There is only one way of interpreting that last sentence, and it is
>not a very pro-Amiga one :-(.
Give me a break, ED is NOT a killer Amiga, it is using Amiga technology.
>I don't wish to dampen people's enthusiasm about Viscorp, but it is
>important that hopes are not unnecessarily raised. The only positive
True.
>thing that I have seen since Toulouse has been the statement from
>Viscorp that they wanted to be a player in the Amiga *computer*
>market. Guess we'll just have to wait and see. And if Viscorp turns
True, but they must realize that the current A1200 And A4000t is
underpowered for todays market and if they want to use off-the-shelf
chips under the TRUE Amiga name computer, without those special Amiga
features, it won't be an Amiga anymore.
>out to be another Escom-like predator who wants its greedy hands on
>our beloved Amy, then there's always Phase 5 :-)
That seems like what they intend to be.. they want the OS and the money
from everything else after they are finished disposing of the AGA chipset
and they move onto their 3rd and 4th generation ED boxes.
As for PhaseV.. I don't know about them yet.. will they obtain licenses
for the Amiga Copper useage etc? What Amiga specifics will they be
continuing with or with their new MAC accelarators, are they just
basically going to be producing an offshoot of a PowerMac?
-=*George*=-
George Noel
In article <4rgrr8$2...@coranto.ucs.mun.ca>,
>: at 64 bit) at a low cost, then imagine the set-top boxes they could make
>: then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
>: sure VIScorp could do amazing things with that kind of technology in a
>: set-top box and a computer!
>
>cpus. Then again it depends on what you call 64 bit too.
Parallel CPUs are a whole new ballgame but would be nice. Maybe a
version of AmigaOS could be ported to the BeBox or emulated for
starters. Parallel CPUs come better into effect when the time comes for
single CPU speeds to reach their maximum price/performance ratio or they
simply cannot be pushed anymore.
>Regards,
>
>BM
-=*George*=-
Ronald van Eijck
In article <<Du1wqG.Gyn...@torfree.net>> Donald Dalley writes:
>Very well put, George! I filled in the VIScorp questionaire, with fewer
>words, saying that they should get the OS into 64-bit mode. If 64 bit
>chips/OS were used, the Amiga would definately have an edge in timing and
>technology. This may be just the ticket.
>I'd like to see some discussion re: a complete 64-bit Amiga.
>Donald
And what might that 64-bit Amiga be then?
Sofar there are very limited areas that need capabilities beyond 32 bit
and the addition of some 64bit math routines to the OS would suffice
for most of them.
If we are talking about 64 bit (or better) graphics buswidth that has
nothing todo with a 64 bit Amiga.
C'ya,
* Ronald van Eijck
*
* If you know the enemy and know yourself, you need not fear
* the result of a hundred battles. (Sun Tzu)
George Noel
In article <4rh16m$a...@usenet.ucs.indiana.edu>,
Lynn Winebarger <owin...@nickel.ucs.indiana.edu> wrote:
>I have heard from a local dealer that DEC will take a binary and convert it
>to an alpha binary free of charge for developers who want to change CPUs.
>Of course, I don't know what the truth value of this information is, but
>reliable in the past.
> If this were true, it would make the software side of the conversion
>quite a bit easier.
This sounds like something VIScorp would be interested in.. no money on
their part, another market for the OS and licensing fees. More then
likely a deal would be DEC port AmigaOS to the DEC Alpha in return for
producing their version of VIScorp's set-top-box. AmigaOS would make the
DEC Alpha seem like that much more better of a CPU and perhaps VIScorp
could even use a DEC Alpha workstation as their main server for providing
online services, run on AmigaOS. :)
>Lynn
-=*George*=-
Jeroen T. Vermeulen
[ A dozen or so newsgroups trimmed ]
In article <4rh16m$a...@usenet.ucs.indiana.edu> owin...@nickel.ucs.indiana.edu (Lynn Winebarger) writes:
> I have heard from a local dealer that DEC will take a binary and convert it
> to an alpha binary free of charge for developers who want to change CPUs.
> Of course, I don't know what the truth value of this information is, but
> the dealer does sell DEC Alpha systems (for rendering) and has been fairly
> reliable in the past.
> If this were true, it would make the software side of the conversion
> quite a bit easier.
You're probably referring to DEC's binary conversion programs from VAX (and MIPS
I believe) to Alpha. This is not a magical universal translation service for
any kind of program for any CPU or OS; it's just an effective migration path for
some of their older customers.
> Lynn
--
============================================================================
# Jeroen T. Vermeulen \"How are we doing kid?"/ Yes, we use Amigas. #
#--- j...@xs4all.nl ---\"Oh, same as always."/-- ... --#
#jver...@wi.leidenuniv.nl \ "That bad, huh?" / Got a problem with that? #
Giving a fuck is the nicest present
Jon Carroll
On 5 Jul 1996 18:55:01 GMT, gn...@morgan.ucs.mun.ca (George Noel)
wrote:
>This sounds like something VIScorp would be interested in.. no money on
>their part, another market for the OS and licensing fees. More then
>likely a deal would be DEC port AmigaOS to the DEC Alpha in return for
>producing their version of VIScorp's set-top-box. AmigaOS would make the
>DEC Alpha seem like that much more better of a CPU and perhaps VIScorp
>could even use a DEC Alpha workstation as their main server for providing
>online services, run on AmigaOS. :)
>
>-=*George*=-
>
I think that even with the size changes the Amiga OS on a DEC would
still be an improvement over
Windows-excuse-me-I-need-12-megs-for-my-OS- NT
Jon Carroll, GSG, SFMC
NCOIC, 11th MSG 'The Wyld Boys'
President, Crescent Moon Productions
dra...@infi.net
"They're all dead, they just don't know it yet."
-Eric Draven, THE CROW
singh
In article <4rjn2s$t...@coranto.ucs.mun.ca>, gn...@morgan.ucs.mun.ca
(George Noel) wrote:
Great. VIScorp sound like a bunch of idiots, worse than ESCOM. Maybe it's
time to get a PC afterall. Our poorman's SGI is pretty much dead as a
real computer, at least from the sounds of this. I don't need a settop
box. To hell with them.
--
wavegirl
"Whaddya think, sirs?"
David Corn
On 5 Jul 1996 16:54:07 GMT, dme...@tiac.net (David Meyer) wrote:
>George Noel (gn...@morgan.ucs.mun.ca) wrote:
>:
And so you're willing to assure me that there are NO facets of the
Amiga that are 16-bit, or use 16-bit hardware?
The fact that a PC uses 16 bit ISA buses for, say, a sound card and
modem (which is all they're used for these days) really doesn't strike
me as being a good argument for negating the saying that PC hardware
is 32 bit.
George Noel
In article <Du1wqG.Gyn...@torfree.net>,
>: PPC line that is 64 bit! Intel will be teaming with Hewlett Packard (HP)
>
>: not try to stay ahead by porting AmigaOS to a 64 bit processor and/or
>: using a 64 bit graphics sub-system as standard to stay ahead of the game?
>: If they can make this lowend, then perfect, they are keeping with the
>: Amiga philosphy, if not, make it highend, it will still be an Amiga and
>: would probably bring in new developers and create new niches. The old
>: developers can stick with what they are used to and go to the 32 bit PPC
>: path with hopes of the ONE (1) so far 64 bit CPU and have a choice to
>: also support a 64 bit total Amiga while the new developers develop for
>: the 64 bit total Amiga and if they chose, also develop for the PPC Amiga
>: compatibles once they see how powerful AmigaOS can be.
>
>: at 64 bit) at a low cost, then imagine the set-top boxes they could make
>: then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
>: sure VIScorp could do amazing things with that kind of technology in a
>: set-top box and a computer!
>
>
>Very well put, George! I filled in the VIScorp questionaire, with fewer
>words, saying that they should get the OS into 64-bit mode. If 64 bit
Yes, a 64 bit processor and the OS in 64 bit even if it had to still have
a 32 bit data bus for now until it can fully go 64 bit. Do you still
have a copy of the VIScorp questionaire?
>chips/OS were used, the Amiga would definately have an edge in timing and
>technology. This may be just the ticket.
Exactly. Intel are only now starting to fully take advantage of 32 bit
and the PPC Amigas would be 32 bit, get a start on either a PPC 64 bit
620 Amiga or a DEC Alpha based 64 bit Amiga.
>I'd like to see some discussion re: a complete 64-bit Amiga.
Use AGA and 68020 - 68040 for now for set-top boxes and lowend computers,
060 for highend Amiga. Use funding from selling of computers to fund AGA+ and
AmigaOS+ License AmigaOS out to PIOS for PPC conversion and they put out
a PPC Amiga compatible. Use all licensing profit and sales profit by
VIScorp to do R&D (along with DEC) for a 32/64 bit motherboard, 64 bit
CPU and use PIOS's PPC AmigaOS code to port it to the DEC Alpha or give
DEC the source code now and port it themselves for free, making AmigaOS
64 bit within 2 - 3 years. In the meantime PPC would be the middle
girlfriend. :)
>Donald
>--
>+ Are you experienced? Observant?
>+ Send your subway/bus rider tips to ab...@torfree.net .
-=*George*=-
Kolbjørn Barmen
Michael Kraemer (kra...@clri6j.gsi.de) wrote:
>
> They don't need to. At the rate they are pumping out new x86 generations
> it's not worthwhile to get a replacement. If your '486 breaks, dump it
> and get a '586 instead. Motorola in contrast needs/needed much more time to
> develop their 68k CPUs. Just look how many years they needed for the 68040
> and 68060. You still can buy 68020/30/40 chips and probably also some
> derivative of the 68000. Don't know if intel still makes 8086/80186/80286/80386,
> probably not.
And what is the marked for intels?
68k's you can find in old VAXes, fridges, TVs, Filmcameras, Communication
equipment, jetfighters, spaceshuttles, sattelites, soundstudios, scanners
broadcaststudios, lightcontrollboards, cars, printers, monitors, etc...
They make quality prosessors that need warranty, they also run heavy tests
on them before releasing them on the marked.
We all remember the fdiv-bug in the first intel pentiums, and how Intel
tried to sweep it under the carpet ;)
Bernhard Graf
In article <4rgap9$d...@due.unit.no>:
"Motorola and the m68k (was "Re: VIScorp's possible plans..")"
KB> George Noel (gn...@morgan.ucs.mun.ca) wrote:
KB> > A 68000, 16 BIT CPU
KB>
KB> I have always wondered about this..
KB>
KB> From one of motorlas webservers, http://129.38.233.1/prod/0X0/68000.html :
KB>
KB>
KB> What's the 16 bit in here?
The data bus is 16 bit.
The registers are 32 bit and the command set offers the appropriate
32bit instructions.
Afaik the M68k's ALU is only 16bit. Therefore 32bit operations take more
cycles than with 16bit ops.
KB> According to Motorola them selves, all the 68k family is 32-Bit.
According to MS, Win95 is 32bit. :)
Ciao
--
Bernhard Graf
mail gr...@jojo.IN-Berlin.DE
graf...@sp.zrz.TU-Berlin.DE
Byron Montgomerie
singh (si...@wwa.com) wrote:
: Great. VIScorp sound like a bunch of idiots, worse than ESCOM. Maybe it's
: time to get a PC afterall. Our poorman's SGI is pretty much dead as a
: real computer, at least from the sounds of this. I don't need a settop
: box. To hell with them.
Now, now, no need to dive off the deep end. :) I think that all the
concerns that might be raised have been, let's just wait and see what
VISCorp has to say about all of it. You have to realise that all those
amiga people from the past have to get caught up with the way things are
with the amiga today. If they are not too stuck in their ways or off in
some high tech dream world. :) It would be nice if they didn't just think
of how to make a few bucks right away and actually planned long turn, it
would be a nice change.
Before trying to get a good deal for a lot of unpopular cpus like the dec
alpha they should look into Pentium chips or their clones. That is if the
target is a low cost machine, PCs have a lot of cool parts just hooked up in
a non optimal way because of backwards compatibility. Why not?
Regards,
BM
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >I'd like to see some discussion re: a complete 64-bit Amiga.
: Use AGA and 68020 - 68040 for now for set-top boxes and lowend computers,
: 060 for highend Amiga. Use funding from selling of computers to fund AGA+ and
Too late for that. 060 as a possible for the low end, although who would
buy it know the os was going PPC?
: AmigaOS+ License AmigaOS out to PIOS for PPC conversion and they put out
: a PPC Amiga compatible. Use all licensing profit and sales profit by
: VIScorp to do R&D (along with DEC) for a 32/64 bit motherboard, 64 bit
: CPU and use PIOS's PPC AmigaOS code to port it to the DEC Alpha or give
: DEC the source code now and port it themselves for free, making AmigaOS
: 64 bit within 2 - 3 years. In the meantime PPC would be the middle
: girlfriend. :)
Too little, too late, where do you realistically see the PC in 2-3 years?
(No fantasies of them suddenly disappearing either) The playstation is
good because it has 3D co-processing, not because it is 64 bit, that is
largely a marketing gimmick. There is no sense in planning to catch up to
the PC as they are today in 2-3 years.
Regards,
BM
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >Personally, I would rather they went with parrallel cpus before using 64 bit
: >cpus. Then again it depends on what you call 64 bit too.
: Parallel CPUs are a whole new ballgame but would be nice. Maybe a
: version of AmigaOS could be ported to the BeBox or emulated for
: starters. Parallel CPUs come better into effect when the time comes for
: single CPU speeds to reach their maximum price/performance ratio or they
: simply cannot be pushed anymore.
Not the right time for 64 bit ram that is for sure, just when 32 bit ram
becomes dirt cheap you want to use 64 bit ram. :) If you mean a 32 bit data
path with 64 bit registers that is a non issue anyway.
Regards,
BM
Kolbjørn Barmen
David Corn (dc...@paradise.pplnet.com) wrote:
>
> And so you're willing to assure me that there are NO facets of the
> Amiga that are 16-bit, or use 16-bit hardware?
As I mentioned earlier, the filesystem has 16 flag bits :)
The 68000 has a 16bit bus, the ZorroII-bus was made for 68000,
ZorroII is 16bit.
> The fact that a PC uses 16 bit ISA buses for, say, a sound card and
> modem (which is all they're used for these days) really doesn't strike
> me as being a good argument for negating the saying that PC hardware
> is 32 bit.
The ZorroIII-bus is 32bit, my gfx card uses internal 64bit,
my ethernetcards are 16bit (they are zorroII)
Brian Turmelle
Jon Carroll wrote:
>
> I think that even with the size changes the Amiga OS on a DEC would
> still be an improvement over
> Windows-excuse-me-I-need-12-megs-for-my-OS- NT
>
Hmm. I can get 32MB in 2 60ns 72pin 16MB SIMMS for a whopping $260.
Oh no, that's like less than half a weeks paycheck. <g>
Julian Regel
gn...@morgan.ucs.mun.ca (George Noel) wrote:
erPC
>Amiga compatibles would suit there nicely. The DEC Alpha Amigas could be
>aimed at the highend for now unless DEC comes up with a stripped down
>version of their processors/cost reduce them to be used in a home market
Hmm, this sounds like the StrongArm processor developed by Arm and
Digital. The early StrongArms run at 150Mhz I believe...
>joystick ports :) ) so why couldn't VIScorp have a DEC Alpha core graphic
>chipset (Carl Sassenrath did say they were talking to R.J. Mical [one of
>the orginal Amiga chipset designers] and that some ideas were tossed
>around) and still use a PPC CPU?
That would be pretty cool.
>Right now a majority of home computers are running on 32 bit hardware, so
>are the game consoles (Atari [debateable] and Nintendo [soon] aside).
Atari is 64bit. That's not debatable (well, it shouldn't be ;-)
>If they could get this 64 bit technology (don't forget DEC Alpha's START
>at 64 bit) at a low cost, then imagine the set-top boxes they could make
>then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
>sure VIScorp could do amazing things with that kind of technology in a
>set-top box and a computer!
I like a lot of the things you've said here. I'll be watching this
thread to see what happens....
Trepanation
: : VIScorp to do R&D (along with DEC) for a 32/64 bit motherboard, 64 bit
: : CPU and use PIOS's PPC AmigaOS code to port it to the DEC Alpha or give
: : DEC the source code now and port it themselves for free, making AmigaOS
: : 64 bit within 2 - 3 years. In the meantime PPC would be the middle
: : girlfriend. :)
:
: Too little, too late, where do you realistically see the PC in 2-3 years?
: (No fantasies of them suddenly disappearing either) The playstation is
: good because it has 3D co-processing, not because it is 64 bit, that is
: largely a marketing gimmick. There is no sense in planning to catch up to
: the PC as they are today in 2-3 years.
:
: Regards,
:
: BM
As he stated, PCs are only now really taking advantage of 32-bit OS and
hardware. If, in 2-3 years, the AmigaOS and hardware optimized 64-bit
technology, that would be competitive with what the PC is likely to
evolve to, not what they are today; i.e. 64-bit *then* isn't competing
with 32-bit *now*.
Moo...
Muraii/\Flabberghasted
--
"Procrustes in modern dress, the nuclear scientist will prepare the bed
on which mankind must lie; and if mankind doesn't fit--well, that will be
just too bad for mankind." Dr. Aldous Huxley, (c) 1946
Trepanation
Brian Turmelle (bri...@concentric.net) wrote:
Still, the current affordability of RAM needn't be a justification for
writing a bloated OS.
Moo...
Muraii/\Slamp
Trepanation
Julian Regel (jre...@melech.demon.co.uk) wrote:
: gn...@morgan.ucs.mun.ca (George Noel) wrote:
:
:
: >Right now a majority of home computers are running on 32 bit hardware, so
:
: Atari is 64bit. That's not debatable (well, it shouldn't be ;-)
What I had heard is that the Atari Jaguar used two 32-bit CPUs, and that
really isn't a *true* 64-bit machine.
Moo...
Muraii/\Jihad
Jean Klos
>
>
> (Including EC, LC, and HC)
>
> NOTE: Motorola no longer makes the MC68000 only it's derivatives.
>
> * 32-Bit Data and Address Registers
> * Addressing Range
> * Program Counter
> * 56 Powerful Instructions
> * Operations on Five Main Data Types
> * Memory-Mapped Input/Output (I/O)
> * 14 Addressing Modes
>
> What's the 16 bit in here?
>
>
>
> I can imagine intel give 5 years warranty on their products... NOT!
>
> Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
> ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
The 680x0 all have 16 bit instruction words; hence the "16/32 bit"
designation.
As far as all of the other posts are concerned, some ask why Motorola
still sells older 680x0, whereas Intel doesn't sell many 80x86. Simple.
How functional is an old Amiga or an old Mac? Not very functional, but
not completely dead. Most software will still run on old 68000s. Try
running anything on an 8086 or a 286. Intel blew up many bridges and
still have a crappy 8 bit instruction core in every Pentium made.
Jean Klos
Amiga 1200, 50 mhz '060, 66 meg, 1 gig Seagate Hawk DMA SCSI-2
Want to see Mac apps multitask?
George Noel
In article <4rkcjr$t...@coranto.ucs.mun.ca>,
>
>: Use AGA and 68020 - 68040 for now for set-top boxes and lowend computers,
>: 060 for highend Amiga. Use funding from selling of computers to fund AGA+ and
>
>Too late for that. 060 as a possible for the low end, although who would
>buy it know the os was going PPC?
Well the 68020-040, namely 030 and 040 for lowend computers might be JUST
pushing its time, lowend 040 MACS are still selling. For set-top boxes,
this is fine for now. AGA+ won't do wonders for the Amiga computer but it
will make the set-top-box better or now. Lowend 060 is out of the question.
>: AmigaOS+ License AmigaOS out to PIOS for PPC conversion and they put out
>: a PPC Amiga compatible. Use all licensing profit and sales profit by
>: VIScorp to do R&D (along with DEC) for a 32/64 bit motherboard, 64 bit
>: CPU and use PIOS's PPC AmigaOS code to port it to the DEC Alpha or give
>: DEC the source code now and port it themselves for free, making AmigaOS
>: 64 bit within 2 - 3 years. In the meantime PPC would be the middle
>: girlfriend. :)
>
>Too little, too late, where do you realistically see the PC in 2-3 years?
>(No fantasies of them suddenly disappearing either) The playstation is
I see that as the only logical step to take unless you go with
multiprocessing as the other course. PCs time have come.. with the such
a large downfall in buying of PCs this year due to the large system
resources you need to upgrade to just to run Windows or the latest
games. They are only now getting into 32 bit so how can you say getting
a lead into 64 is too little, too late? PCs are still using a CISC
design, by next year Amiga compatibles will be using a RISC design and
once the HP/Intel RISC chip comes out, Intel do not guarantee that old
programs will run under it. This means PC users either start aknew with
the new processor or check out a different system.. either way, most of
their old software will not work.
>good because it has 3D co-processing, not because it is 64 bit, that is
>largely a marketing gimmick. There is no sense in planning to catch up to
The Playstation is overall 32 bit. There is some debate as to whether
Atari's Jaguar is really the first 64 bit game console, besides that,
the first true 64 bit game console will be the Nintendo 64 introduced in
North America in September.
>the PC as they are today in 2-3 years.
Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
Alphas start at around 300 Mhz from what I can remember..NOW. PPC 620 is
rumoured to achieve 300 MHz after tweaking (around 133 Mhz right now I
think) and future PPC hybrids from a different company (to compete with
IBM/Motorola) rumoured to achieve 500 - 700 Mhz. The clones have
marketing, Microsoft and the business sector behind them but like I said,
their time is coming. If other companies started today, in a couple of
years they could have a real serious alternative on their hands, WITH the
upperhand.
>Regards,
>
>BM
-=*George*=-
B.Mikic
Brian Turmelle <bri...@concentric.net> wrote:
>Jon Carroll wrote:
>>
>> I think that even with the size changes the Amiga OS on a DEC would
>> still be an improvement over
>> Windows-excuse-me-I-need-12-megs-for-my-OS- NT
>>
>Hmm. I can get 32MB in 2 60ns 72pin 16MB SIMMS for a whopping $260.
>Oh no, that's like less than half a weeks paycheck. <g>
see my signature...
----- -----
, "A Wintel user solves a problem
B.Mikic by throwing money on it."
----- -----
John Sheehy
ko...@nvg.unit.no (Kolbjørn Barmen) writes:
>We all remember the fdiv-bug in the first intel pentiums, and how Intel
>tried to sweep it under the carpet ;)
Move16 - the motorola 040 instruction that didn't work right. It was a
new instruction, so it didn't cause any backward compatibility problems.
To keep it from crashing the machine, you have to pad it with NOPs, and
of course you need exclusive possession of the CPU to do that.
<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <jsh...@ix.netcom.com>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
David Corn
On 6 Jul 1996 22:09:23 GMT, gn...@morgan.ucs.mun.ca (George Noel)
wrote:
>Well the 68020-040, namely 030 and 040 for lowend computers might be JUST
>pushing its time, lowend 040 MACS are still selling. For set-top boxes,
They are? Not in the USA. You -might- be able to find an old
Performa 040 on a rack somewhere, and you might be able to find an 040
in an old portable somewhere, but Apple, for the most part, has gone
PPC-only.
>this is fine for now. AGA+ won't do wonders for the Amiga computer but it
>will make the set-top-box better or now. Lowend 060 is out of the question.
The 060 is the low end. It's where the Amiga should start.
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >Too late for that. 060 as a possible for the low end, although who would
: >buy it know the os was going PPC?
: Well the 68020-040, namely 030 and 040 for lowend computers might be JUST
: pushing its time, lowend 040 MACS are still selling. For set-top boxes,
: this is fine for now. AGA+ won't do wonders for the Amiga computer but it
: will make the set-top-box better or now. Lowend 060 is out of the question.
Unless you have VISCorp stock, who cares about their set top box? So you
think an 030 or 040 (expensive) aga+ amiga is the appropriate next step in
the amiga line? If they sell them at $300 or so perhaps, which would be
as realistic as the idea that the PC will suddenly disappear and eveyone
will start to use amigas instead.
: >Too little, too late, where do you realistically see the PC in 2-3 years?
: >(No fantasies of them suddenly disappearing either) The playstation is
: I see that as the only logical step to take unless you go with
: multiprocessing as the other course. PCs time have come.. with the such
: a large downfall in buying of PCs this year due to the large system
: resources you need to upgrade to just to run Windows or the latest
: games. They are only now getting into 32 bit so how can you say getting
Wishfull thinking.
: a lead into 64 is too little, too late? PCs are still using a CISC
64 bit data buses are not magic, what that means is that a 33mhz motherboard
will perform like a 66mhz one. 64 bit or a proper cache, which is more
practical? You going to buy a motherboard that is twice as expensive?
: design, by next year Amiga compatibles will be using a RISC design and
: once the HP/Intel RISC chip comes out, Intel do not guarantee that old
: programs will run under it. This means PC users either start aknew with
: the new processor or check out a different system.. either way, most of
: their old software will not work.
Who care what Intel does, if they drop the ball there then Cyrix or AMD will
take up the slack.
: >good because it has 3D co-processing, not because it is 64 bit, that is
: >largely a marketing gimmick. There is no sense in planning to catch up to
: >the PC as they are today in 2-3 years.
: Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
For this month.
: Alphas start at around 300 Mhz from what I can remember..NOW. PPC 620 is
: rumoured to achieve 300 MHz after tweaking (around 133 Mhz right now I
: think) and future PPC hybrids from a different company (to compete with
: IBM/Motorola) rumoured to achieve 500 - 700 Mhz. The clones have
vapour. Cpu clock rates don't tell the entire story anyway.
: marketing, Microsoft and the business sector behind them but like I said,
: their time is coming. If other companies started today, in a couple of
: years they could have a real serious alternative on their hands, WITH the
: upperhand.
I guess you believe in the tooth fairy as well? :)
Regards,
BM
Byron Montgomerie
Trepanation (mur...@nmia.com) wrote:
: : Too little, too late, where do you realistically see the PC in 2-3 years?
: : (No fantasies of them suddenly disappearing either) The playstation is
: : largely a marketing gimmick. There is no sense in planning to catch up to
: : the PC as they are today in 2-3 years.
: As he stated, PCs are only now really taking advantage of 32-bit OS and
: hardware. If, in 2-3 years, the AmigaOS and hardware optimized 64-bit
: technology, that would be competitive with what the PC is likely to
: evolve to, not what they are today; i.e. 64-bit *then* isn't competing
: with 32-bit *now*.
So the amiga should have a 64 bit (data bus) motherboard at what clock
rating? You willing to pay for that? If it were cost effective the PC
would have it by now. There are more important priorities right now IMO.
In the last few months the Pentium has gone from 133Mhz as the top speed to
200Mhz+, you willing to bet that 64 bit would be enough to compete in 2-3
years time? Will the dec alpha even be a supported cpu then?
Regards,
BM
George Noel
In article <4rmavm$v...@hume.nmia.com>, Trepanation <mur...@nmia.com> wrote:
>Byron Montgomerie (bmon...@morgan.ucs.mun.ca) wrote:
>: George Noel (gn...@morgan.ucs.mun.ca) wrote:
>:
>: : VIScorp to do R&D (along with DEC) for a 32/64 bit motherboard, 64 bit
>: : CPU and use PIOS's PPC AmigaOS code to port it to the DEC Alpha or give
>: : DEC the source code now and port it themselves for free, making AmigaOS
>: : 64 bit within 2 - 3 years. In the meantime PPC would be the middle
>: : girlfriend. :)
>: Too little, too late, where do you realistically see the PC in 2-3 years?
>: (No fantasies of them suddenly disappearing either) The playstation is
>: good because it has 3D co-processing, not because it is 64 bit, that is
>: largely a marketing gimmick. There is no sense in planning to catch up to
>: the PC as they are today in 2-3 years.
>:
>:
>
>As he stated, PCs are only now really taking advantage of 32-bit OS and
>hardware. If, in 2-3 years, the AmigaOS and hardware optimized 64-bit
The Pentium Pro is the first Intel CPU to practically be optimised fully
for 32 bit and Windows isn't entirely 32 bit with '95 (even though Bill
Gates said it would be). This made Intel's Pentium Pro look REAL good!
:) The 16 bit parts in Windows actually slowed things down on the Pentium
Pro! :) Now Bill is saying Windows will be entirely 32 bit by '97! :)
>technology, that would be competitive with what the PC is likely to
>evolve to, not what they are today; i.e. 64-bit *then* isn't competing
>with 32-bit *now*.
True and I don't think Intel are going to be pumping out too many
versions of their CPU for the next while. The jump from the 60, 75, 90,
133, 166 and 200 MHz CPUs within 14 months really caused a lot of people
to sit back and say WHOAA! Slow down! And companies to write off massive
amounts of hardware making them lose money when no one was buying the
newer CPUs. No wonder PCs are so cheap.
>Moo...
>
>
>Muraii/\Flabberghasted
>--
>"Procrustes in modern dress, the nuclear scientist will prepare the bed
>on which mankind must lie; and if mankind doesn't fit--well, that will be
>just too bad for mankind." Dr. Aldous Huxley, (c) 1946
-=*George*=-
George Noel
In article <4rmb5c$v...@hume.nmia.com>, Trepanation <mur...@nmia.com> wrote:
>Julian Regel (jre...@melech.demon.co.uk) wrote:
>: gn...@morgan.ucs.mun.ca (George Noel) wrote:
>:
>:
>: >Right now a majority of home computers are running on 32 bit hardware, so
>: >are the game consoles (Atari [debateable] and Nintendo [soon] aside).
>:
>: Atari is 64bit. That's not debatable (well, it shouldn't be ;-)
>:
>
>What I had heard is that the Atari Jaguar used two 32-bit CPUs, and that
>really isn't a *true* 64-bit machine.
But it has 2 64 bit graphic coprocessors and a 64 bit memory address space.
>Moo...
>
>
>Muraii/\Jihad
George Noel
In article <31df1c67....@news.onramp.net>,
David Corn <dc...@paradise.pplnet.com> wrote:
>On 6 Jul 1996 22:09:23 GMT, gn...@morgan.ucs.mun.ca (George Noel)
>wrote:
>
>>pushing its time, lowend 040 MACS are still selling. For set-top boxes,
>
>Performa 040 on a rack somewhere, and you might be able to find an 040
>in an old portable somewhere, but Apple, for the most part, has gone
>PPC-only.
Yeah, had to write off the 68k line to save its a$$. I was talking about
MAC compatibles really.
>>this is fine for now. AGA+ won't do wonders for the Amiga computer but it
>>will make the set-top-box better or now. Lowend 060 is out of the question.
>
>The 060 is the low end. It's where the Amiga should start.
Yeah right.. have you seen the price for the processor alone? PPC chips
are cheaper then the 060.
-=*George*=-
George Noel
In article <4rn5ir$c...@coranto.ucs.mun.ca>,
>: >buy it know the os was going PPC?
>
>: pushing its time, lowend 040 MACS are still selling. For set-top boxes,
>
If they plan on selling millions of them like they say they are going to,
then that will bring back developers to the Amiga/ED.
>think an 030 or 040 (expensive) aga+ amiga is the appropriate next step in
>the amiga line? If they sell them at $300 or so perhaps, which would be
For VIScorp, within the next 1 year? Yeah. Yes, for a low price point
also.. anywhere from $300 - $500 US or so.
>as realistic as the idea that the PC will suddenly disappear and eveyone
>will start to use amigas instead.
Nah, that will take a few extra years! :)
>: >Too little, too late, where do you realistically see the PC in 2-3 years?
>: >(No fantasies of them suddenly disappearing either) The playstation is
>
>: I see that as the only logical step to take unless you go with
>: multiprocessing as the other course. PCs time have come.. with the such
>: a large downfall in buying of PCs this year due to the large system
>: resources you need to upgrade to just to run Windows or the latest
>: games. They are only now getting into 32 bit so how can you say getting
>
>Wishfull thinking.
It is starting to unfold this year.. heck even Microsoft put off putting
a Windows '96 out due to lack of expected Windows '95 sales.
>: design, by next year Amiga compatibles will be using a RISC design and
>: once the HP/Intel RISC chip comes out, Intel do not guarantee that old
>: programs will run under it. This means PC users either start aknew with
>: the new processor or check out a different system.. either way, most of
>: their old software will not work.
>
>Who care what Intel does, if they drop the ball there then Cyrix or AMD will
>take up the slack.
But their CPUs are crap compared to Intel and even Intel is not the best
CPU makers out there.
>: Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
>
>For this month.
No, I expect the whole industry is going to slow down now due to
saturation and poor sales. It has started in the memory production area
and I doubt Intel are going to want to piss off their buyers again this
soon by putting out another optimised version of their CPU this soon.
>: marketing, Microsoft and the business sector behind them but like I said,
>: their time is coming. If other companies started today, in a couple of
>: years they could have a real serious alternative on their hands, WITH the
>: upperhand.
>
>I guess you believe in the tooth fairy as well? :)
No, that is for PC users, if Bill Gates says so. :)
>Regards,
>
>BM
-=*George*=-
George Noel
In article <4rn695$c...@coranto.ucs.mun.ca>,
Byron Montgomerie <bmon...@morgan.ucs.mun.ca> wrote:
>Trepanation (mur...@nmia.com) wrote:
>: As he stated, PCs are only now really taking advantage of 32-bit OS and
>: hardware. If, in 2-3 years, the AmigaOS and hardware optimized 64-bit
>: with 32-bit *now*.
>
>rating? You willing to pay for that? If it were cost effective the PC
If DEC doesn't have a cost reduced version of their Alphas, then at least
300 Mhz. :) Depending on the price and if it is meant as a highend or
lowend. Like I said, they could be doing it to offer their customers
servers to run online services off of, with the servers running AmigaOS.
Eventually a long way down the road, people could have this type of
computer as the main computer in their home, with seperate lowend Amigas
in practically each room networked with the main server. :) Yes, I am
dreaming now. :)
>would have it by now. There are more important priorities right now IMO.
PCs are not God, they have good marketing, had the business community
behind them and were able to increase their clock speeds moreso to make
up for the drawbacks in the PC architecture. Of course the competition
helped out also. They wouldn't have it by now as it would totally break
their upgradeability path, making them lose a lot of customers. Even the
Pentiums are still linked to the original 8086.
>In the last few months the Pentium has gone from 133Mhz as the top speed to
>200Mhz+, you willing to bet that 64 bit would be enough to compete in 2-3
>years time? Will the dec alpha even be a supported cpu then?
Compete - yes. And the quickness in upgrading their CPUs has hurt
Intel. As for DEC being supported then.. if they don't do something like
this soon, perhaps not.
>Regards,
>
>BM
-=*George*=-
Brian Turmelle
Trepanation wrote:
>
> Brian Turmelle (bri...@concentric.net) wrote:
> : Jon Carroll wrote:
> : >
> : > I think that even with the size changes the Amiga OS on a DEC would
> : > still be an improvement over
> : > Windows-excuse-me-I-need-12-megs-for-my-OS- NT
> : >
> :
> : Hmm. I can get 32MB in 2 60ns 72pin 16MB SIMMS for a whopping $260.
> : Oh no, that's like less than half a weeks paycheck. <g>
>
> writing a bloated OS.
What's so bloated about it. Do you know what NT was designed for and what it
can do? If you did then you wouldn't squibble over 12MB RAM. NT is not
designed for a single user. It is a multiuser, preemtive multitasking
multithreading powerhouse. It even supports multiple processors. We
have a Compaq here at work that has 4 P133's in it, with 256MB RAM. Windows NT
runs on x86, DEC Alpha and PowerPC processors. Of course it wasn't fully
optimized for each processor. If AmigaOS gets anywhere near as powerful as NT
then we'll see who has a bloated OS.
Brian Turmelle
Trepanation wrote:
>
> Julian Regel (jre...@melech.demon.co.uk) wrote:
> : gn...@morgan.ucs.mun.ca (George Noel) wrote:
> :
> :
> : >Right now a majority of home computers are running on 32 bit hardware, so
> : >are the game consoles (Atari [debateable] and Nintendo [soon] aside).
> :
> : Atari is 64bit. That's not debatable (well, it shouldn't be ;-)
> :
>
> What I had heard is that the Atari Jaguar used two 32-bit CPUs, and that
> really isn't a *true* 64-bit machine.
>
I believe it had a 68000 and 2 32 bit CPU's, and a 64bit BUS. The 32bit CPU's did
have a couple 64bit operations though. Which is why they were able to get away
with calling it a 64bit machine. It was able to move up to 64bits at a time.
Brian Turmelle
Jean Klos wrote:
>
> As far as all of the other posts are concerned, some ask why Motorola
> still sells older 680x0, whereas Intel doesn't sell many 80x86. Simple.
> How functional is an old Amiga or an old Mac? Not very functional, but
> not completely dead. Most software will still run on old 68000s. Try
> running anything on an 8086 or a 286. Intel blew up many bridges and
> still have a crappy 8 bit instruction core in every Pentium made.
>
That's not why Intel doesn't sell that many of those processors. It's
because the Intel clones sweep up the lowend market with their own versions
of them. AMD, Cyrix, and Nexgen sell them instead. While Intel focuses on
creating faster and better processors. Which is where the money is.
I was part of a satellite design team that used a radiation hardened 8086 as
it's core. Does Motorolla make radiation hardened 68000's?
David Corn
On 7 Jul 1996 04:44:02 GMT, gn...@morgan.ucs.mun.ca (George Noel)
wrote:
>Yeah, had to write off the 68k line to save its a$$. I was talking about
>MAC compatibles really.
Power Computing is the big one, and they've always been all PPC,
haven't they?
>>The 060 is the low end. It's where the Amiga should start.
>
>Yeah right.. have you seen the price for the processor alone? PPC chips
>are cheaper then the 060.
Great - then go PPC. So who's going to write this new OS, and do you
think it will be finished by the year 2000?
Ronald van Eijck
In article <<4rmb24$v...@hume.nmia.com>> Trepanation writes:
>Brian Turmelle (bri...@concentric.net) wrote:
>: Jon Carroll wrote:
>: >
>: > I think that even with the size changes the Amiga OS on a DEC would
>: > still be an improvement over
>: > Windows-excuse-me-I-need-12-megs-for-my-OS- NT
>: >
>:
>: Hmm. I can get 32MB in 2 60ns 72pin 16MB SIMMS for a whopping $260.
>: Oh no, that's like less than half a weeks paycheck. <g>
>Still, the current affordability of RAM needn't be a justification for
>writing a bloated OS.
No, but it is a good indication that spending many development hours
just to lower the memory requirements for a program by half a megabyte
in not the path to the future. It might be cheaper to just supply that
memory expansion with the software :-)
C'ya,
* Ronald van Eijck
*
* If you know the enemy and know yourself, you need not fear
* the result of a hundred battles. (Sun Tzu)
Brian Turmelle
George Noel wrote:
>
> I see that as the only logical step to take unless you go with
> multiprocessing as the other course. PCs time have come.. with the such
> a large downfall in buying of PCs this year due to the large system
> resources you need to upgrade to just to run Windows or the latest
> games. They are only now getting into 32 bit so how can you say getting
> once the HP/Intel RISC chip comes out, Intel do not guarantee that old
> programs will run under it. This means PC users either start aknew with
> the new processor or check out a different system.. either way, most of
> their old software will not work.
Such a large downfall? Large system resources you need to upgrade? PC's are
still selling very very strong. My fathers 486/33 with 8MB RAM didn't need any
upgrading to run Windows 95. I'll probably upgrade it to a P133+ with 16MB RAM
for Christmas but that won't be very expensive at all($300). My friends 486/66
that's 3 years old runs Windows 95 just fine and most of the newer games. BTW,
there are PC's now that use a RISC design(Cyrix's Pentium clones). P7's are
already in silicone, and will surely be out before AmigaOS gets ported. If
Cyrix can make a RISC Intel clone that's 100% compatable I'm sure Intel can
too. Intel has bent over backwards to remain fully compatable. I highly doubt
most of it will not run.
> The Playstation is overall 32 bit. There is some debate as to whether
> Atari's Jaguar is really the first 64 bit game console, besides that,
> the first true 64 bit game console will be the Nintendo 64 introduced in
> North America in September.
Which is already in Japan now.
> >the PC as they are today in 2-3 years.
>
> Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
> Alphas start at around 300 Mhz from what I can remember..NOW. PPC 620 is
> rumoured to achieve 300 MHz after tweaking (around 133 Mhz right now I
> think) and future PPC hybrids from a different company (to compete with
> IBM/Motorola) rumoured to achieve 500 - 700 Mhz. The clones have
> marketing, Microsoft and the business sector behind them but like I said,
> their time is coming. If other companies started today, in a couple of
> years they could have a real serious alternative on their hands, WITH the
> upperhand.
And as we all know, it doesn't matter how well designed a system or processor
is. It depends on the software. Also, megahertz alone doesn't mean a thing.
It's how many instructions it can process in a CPU cycle that matters. The only
real serious alternative is a company like Cyrix (IBM fabricates them,
competing with themselves) that has a pin compatable Pentium(not Pro) processor
that runs 25% faster at the same Mhz. PPC has a good design but it isn't fast
enough to compete with Intel. They don't have the resources, even with IBM,
Motorolla and Apple combining together. There's a company that's got some
new pattents for BiCMOS fabricating that's supposed to double current PPC
speeds, but I'm sure Intel can get into that too. Intel is adding more things
to keep the Pentium line ahead. Like the 86 MMX (Multimedia Extensions) to the
X86 instruction set.
Of course processors will get faster as production facilities perfect their
operations. The original pentium started at 60Mhz and has now tweaked to
200Mhz. The Pentium Pro started at 150Mhz and is also supposed to achieve
300Mhz-400Mhz with die size reductions and "tweaking". I forget what the P7 is
supposed to start at. I'm curious what Cyrix will come out with next.
David Corn
On 7 Jul 1996 06:28:56 GMT, gn...@morgan.ucs.mun.ca (George Noel)
wrote:
>It is starting to unfold this year.. heck even Microsoft put off putting
>a Windows '96 out due to lack of expected Windows '95 sales.
Did they? (Source, please?) From what I hear MS will have it in OEM
hands within the next few months. It's been in beta and user review
for months now.
>>Who care what Intel does, if they drop the ball there then Cyrix or AMD will
>>take up the slack.
>
>But their CPUs are crap compared to Intel and even Intel is not the best
>CPU makers out there.
Really? I think Cyrix does a very good job with CPUs - their 150+
models are $260 at local shops, with a good motherboard and 256k
cache. I think it's great - finally some viable competition for
Intel. They've outperformed Intel base systems for quite some time
now, and they've been faster to market with chip technology from the
Pentium 166 onward.
>No, I expect the whole industry is going to slow down now due to
>saturation and poor sales. It has started in the memory production area
>and I doubt Intel are going to want to piss off their buyers again this
>soon by putting out another optimised version of their CPU this soon.
:) You obviously haven't heard about the MMX then? :)
bme...@bruce.cs.monash.edu.au
bmon...@morgan.ucs.mun.ca (Byron Montgomerie) writes:
>George Noel (gn...@morgan.ucs.mun.ca) wrote:
>: [PCs] are only now getting into 32 bit so how can you say getting
>: a lead into 64 is too little, too late? PCs are still using a CISC
>64 bit data buses are not magic, what that means is that a 33mhz motherboard
>will perform like a 66mhz one. 64 bit or a proper cache, which is more
>practical? You going to buy a motherboard that is twice as expensive?
Which brings up an important point! Some people here have said that
the new Amiga should have a 64 bit RISC chip, and suggested using a 32
bit databus design for the low end machines. That is balooney!
There is two main problems these days with building a fast processor.
One is how to make it superscalar (i.e. pipelining and multiple execution
units) and actually keep the pipelines from stalling; The other one is
memory bandwidth. The actual width of the internal registers is not really
all that important anymore, especially given the fact that most CPU intensive
stuff happens in the FPU, anyway, which is a different matter altogether.
Now, Intel and DEC understood the memory bandwidth problem --- not only
did they integrate 1st and second level cache with their processors
(96k 2nd level cache on the 21164, 256 or 512k on the PPro), but they
also gave them wide data busses --- wider in fact than the processor
registers (64bit for Pentiums, up to 256 bit for the 21164).
Bernie
--
==============================================================================
Still thinking the Scots should have gone through to the quarterfinals.....
Kolbjørn Barmen
Brian Turmelle (bri...@concentric.net) wrote:
> What's so bloated about it. Do you know what NT was designed for and what it
> can do? If you did then you wouldn't squibble over 12MB RAM. NT is not
> designed for a single user. It is a multiuser, preemtive multitasking
> multithreading powerhouse. It even supports multiple processors. We
> have a Compaq here at work that has 4 P133's in it, with 256MB RAM.
And how many users in at once?
> Windows NT runs on x86, DEC Alpha and PowerPC processors.
*blurp*
Linux/i386
Linux/alpha
Linux/m68k (amiga, atari, mvme)
mkLinux/PPC for Apple Macintosh
Linux/PPC
Linux/SPARC
Linux/MIPS
Linux/ARM (acorn)
Linux/SMP
NetBSD-alpha
NetBSD-m68k (mac, atari. amiga. mvme, sharp)
NetBSD-arm32 (acorn)
NetBSD-hp300 (HP 9000/300 series)
NetBSD-i386 (family PC)
NetBSD-pc532 (PC532)
NetBSD-pica (Acer Pica)
NetBSD-pmax (DECstation)
NetBSD-sparc (Sun SPARC series)
NetBSD-sun3 (Sun 3)
NetBSD-vax (DEC VAX)
And they're for free!!
"Gå hjem og vogg" as we say it in Norway.
> Of course it wasn't fully optimized for each processor.
> If AmigaOS gets anywhere near as powerful as NT then we'll see who has
> a bloated OS.
"Gosh.. we're real impressed down here"
Apropos multiple processors.. NT shows a flattening curve in performance
when you add more CPUs, after 8 CPUs the system spends just as much
more CPU-time to to administrate the signalprosessing as the the extra
CPU is capable of doing. _Real_ multiple OSes show a stright line in
performance when adding more CPUs, even after CPU number 8.
Source? Dont remember, some OpenWorld some months ago or so I think...
Whats this posting's got to do with the amiga? Well, for one the above
unixes, Linux and NetBSD, are running on amigas per today, and they're
getting better from day to day. I recommend people with HD-space, MMU and
some amount of patients to check them out. At least it's quite educative.
Second, it's the above OSes AmigaOS people should look at when creating
the core of the "new" AmigaOS, not some bloated unix-wanna-be OS from some
dodgy softewarehouse infameous for it's buggy software, and aggressiv
marketing techniques.
Make AmigaOS a fully loaded bsd for PPC (and perhaps even DEC Alpha), and
I would have no doubt on what platform I will choose in the future.
Remember unix doesnt neccesserly imply user-unfriendlyness,
Thats all up to the developers.
Kolbjørn Barmen
Brian Turmelle (bri...@concentric.net) wrote:
> Such a large downfall? Large system resources you need to upgrade? PC's are
> still selling very very strong. My fathers 486/33 with 8MB RAM didn't need
> any upgrading to run Windows 95. I'll probably upgrade it to a P133+
> with 16MB RAM for Christmas but that won't be very expensive at
> all($300). My friends 486/66 that's 3 years old runs Windows 95 just
> fine and most of the newer games. BTW, there are PC's now that use a
> RISC design(Cyrix's Pentium clones). P7's are already in silicone, and
> will surely be out before AmigaOS gets ported. If Cyrix can make a RISC
> Intel clone that's 100% compatable I'm sure Intel can too. Intel has
> bent over backwards to remain fully compatable. I highly doubt most of
> it will not run.
Sounds like inchip hardware 386-emulation, tsk tsk..
> > The Playstation is overall 32 bit. There is some debate as to whether
> > Atari's Jaguar is really the first 64 bit game console, besides that,
> > the first true 64 bit game console will be the Nintendo 64 introduced in
> > North America in September.
>
> Which is already in Japan now.
September?? =)
> And as we all know, it doesn't matter how well designed a system or
> processor is. It depends on the software.
No kidding, Sherlock!
> Also, megahertz alone doesn't mean a thing.
Mmm..
> It's how many instructions it can process in a CPU cycle that matters.
MIPS? Also known as BOGOMIPS? ;)
> The only real serious alternative is a company like Cyrix (IBM
> fabricates them, competing with themselves) that has a pin compatable
> Pentium(not Pro) processor that runs 25% faster at the same Mhz. PPC
> has a good design but it isn't fast enough to compete with Intel. They
> don't have the resources, even with IBM, Motorolla and Apple combining
> together. There's a company that's got some new pattents for BiCMOS
> fabricating that's supposed to double current PPC speeds, but I'm sure
> Intel can get into that too.
I have a theory they've already done so. The fastes pentium today do not
handle much overclocking before saying bye bye, not that this really
indicates anything..
> Intel is adding more things to keep the Pentium line ahead. Like the 86
> MMX (Multimedia Extensions) to the X86 instruction set.
Hm.. also there is HP's PA-8000 which outraces the DEC Alpha with less
than half clockspeed, and also the forthcomming PA-9000, which is meant to
be a Intel/HP hybrid, with full i386 compatibility.
> Of course processors will get faster as production facilities perfect
> their operations. The original pentium started at 60Mhz and has now
> tweaked to 200Mhz. The Pentium Pro started at 150Mhz and is also
> supposed to achieve 300Mhz-400Mhz with die size reductions and
> "tweaking". I forget what the P7 is supposed to start at. I'm curious
> what Cyrix will come out with next.
And the HP-PA8000@180MHz outruns a DEC-Alpha@400MHz, times are turbulent,
anything can happen.
Kolbjørn Barmen
Ronald van Eijck (ron...@noblehouse.xs4all.nl) wrote:
> No, but it is a good indication that spending many development hours
> just to lower the memory requirements for a program by half a megabyte
> in not the path to the future. It might be cheaper to just supply that
> memory expansion with the software :-)
Take a look at netscape, they use this technique, adding more and more
features, without doing any optimalization.
Notice the tendency of warp from v1.12 up to 3.x-Gold of Navigator?
David Corn
On 7 Jul 1996 14:27:33 GMT, ko...@nvg.unit.no (Kolbjørn Barmen) wrote:
>
>Apropos multiple processors.. NT shows a flattening curve in performance
>when you add more CPUs, after 8 CPUs the system spends just as much
>more CPU-time to to administrate the signalprosessing as the the extra
>CPU is capable of doing. _Real_ multiple OSes show a stright line in
>performance when adding more CPUs, even after CPU number 8.
I also remember reading this. Now, key question - is this MSs or
Intel's fault? In other words, do you see this same problem with
(from 8 -> 9) processors on an Intel box under an Intel Unix?
>Remember unix doesnt neccesserly imply user-unfriendlyness,
>Thats all up to the developers.
Oh great - and you really think you'll sell with an attitude like
that?
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >Unless you have VISCorp stock, who cares about their set top box? So you
: If they plan on selling millions of them like they say they are going to,
: then that will bring back developers to the Amiga/ED.
ED developers, not amiga developers, there is a difference.
: >think an 030 or 040 (expensive) aga+ amiga is the appropriate next step in
: >the amiga line? If they sell them at $300 or so perhaps, which would be
: For VIScorp, within the next 1 year? Yeah. Yes, for a low price point
: also.. anywhere from $300 - $500 US or so.
Won't happen.
: >as realistic as the idea that the PC will suddenly disappear and eveyone
: >will start to use amigas instead.
: Nah, that will take a few extra years! :)
PCs won't be around forever no, but only through evolution into something
more advanced. There are those that say the dinosaurs didn't all just keel
over and die, they evolved into other creatures like birds. :)
: It is starting to unfold this year.. heck even Microsoft put off putting
: a Windows '96 out due to lack of expected Windows '95 sales.
Yeah right, and people will stop using computers suddenly too. :)
: >Who care what Intel does, if they drop the ball there then Cyrix or AMD will
: >take up the slack.
: But their CPUs are crap compared to Intel and even Intel is not the best
: CPU makers out there.
Chicken lips! :)
: >: Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
: >For this month.
: No, I expect the whole industry is going to slow down now due to
: saturation and poor sales. It has started in the memory production area
: and I doubt Intel are going to want to piss off their buyers again this
: soon by putting out another optimised version of their CPU this soon.
You'll be disappointed then, they have to keep improving or they will lose
market share to Cyrix.
: >I guess you believe in the tooth fairy as well? :)
: No, that is for PC users, if Bill Gates says so. :)
Actually, what willy is doing kind of makes sense, keep pushing out the
compatible windows 95 while at the same time developing the next generation
os WindowsNT. What the amiga needs is something similar, keep improving 3.x
on the one hand while developing a clean os from scratch on the other.
Regards,
BM
Regards,
BM
John Sheehy
dc...@paradise.pplnet.com (David Corn) writes:
>On 6 Jul 1996 22:09:23 GMT, gn...@morgan.ucs.mun.ca (George Noel)
>wrote:
>
>>Well the 68020-040, namely 030 and 040 for lowend computers might be JUST
>>pushing its time, lowend 040 MACS are still selling. For set-top boxes,
>
>They are? Not in the USA. You -might- be able to find an old
>Performa 040 on a rack somewhere, and you might be able to find an 040
>in an old portable somewhere, but Apple, for the most part, has gone
>PPC-only.
>
>>this is fine for now. AGA+ won't do wonders for the Amiga computer but it
>>will make the set-top-box better or now. Lowend 060 is out of the question.
>
>The 060 is the low end. It's where the Amiga should start.
What any possible future Amiga really needs is a motherboard overhaul.
The current MBs available can not take full advantage of 040s and 060s.
Their is bottlenecking galore once you leave the CPU and it's local ram.
<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <jsh...@ix.netcom.com>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Trepanation
Brian Turmelle (bri...@concentric.net) wrote:
: Trepanation wrote:
: >
: > : Jon Carroll wrote:
: > : >
: > : > I think that even with the size changes the Amiga OS on a DEC would
: > : > still be an improvement over
: > : > Windows-excuse-me-I-need-12-megs-for-my-OS- NT
: > : >
: > :
: > : Hmm. I can get 32MB in 2 60ns 72pin 16MB SIMMS for a whopping $260.
: > : Oh no, that's like less than half a weeks paycheck. <g>
: >
: > Still, the current affordability of RAM needn't be a justification for
: > writing a bloated OS.
: What's so bloated about it. Do you know what NT was designed for and what it
: designed for a single user. It is a multiuser, preemtive multitasking
: multithreading powerhouse. It even supports multiple processors. We
: runs on x86, DEC Alpha and PowerPC processors. Of course it wasn't fully
: then we'll see who has a bloated OS.
Well, then this should have been your rebuttal to begin with. What I
responded to was your attempt to justify NT's size by saying that RAM is
cheap. Yeah, so what? Now, though, you explain exactly why it is as
large as it is. Nifty, cool. I'd kill for an Alpha running NT and
Lightwave; all I hear about NT is that, once you get it configured
(assuming you haven't set off a thermonuclear device out of frustration),
it is indeed powerful.
But, I'm sorry, it is bloated. I have a certain faith that just about
anything Microsoft codes could have been done twice as efficiently and
therefore require half the resources.
Moo...
Muraii/\Wendigo
Hans-Joerg Frieden
Michael Kraemer (kra...@clri6j.gsi.de) wrote:
: They don't need to. At the rate they are pumping out new x86 generations
: it's not worthwhile to get a replacement. If your '486 breaks, dump it
: and get a '586 instead. Motorola in contrast needs/needed much more time to
This would mean that the 586 would easily plug into a 486 socket, but it
doesn't. And as far as I know, Intel does not make _any_ 586 at all,
these are Cyrix chips...
Regards, Hans-Joerg.
--
Hans-Joerg Frieden |"Why is it, that the years seem to fly past, while
Schloss-Str. 176 |the night seems to last forever?" - Elminster
54293 Trier |"It's like ten thousand spoons when all you need
hfri...@fix.uni-trier.de | is a knife" - Alanis Morissette
Hans-Joerg Frieden
John Sheehy (jsh...@ix.netcom.com) wrote:
: ko...@nvg.unit.no (Kolbjørn Barmen) writes:
:
: >We all remember the fdiv-bug in the first intel pentiums, and how Intel
: >tried to sweep it under the carpet ;)
:
: Move16 - the motorola 040 instruction that didn't work right. It was a
: new instruction, so it didn't cause any backward compatibility problems.
: To keep it from crashing the machine, you have to pad it with NOPs, and
: of course you need exclusive possession of the CPU to do that.
the number was on a segment boundary. The 68000 had a mask fault that
made some instructions perform in user mode whereasall further
generations mad them supervisor mode instructions. There's mailing lists
with processor bug announcements. The point is the last part of the
sentence: Intel tried to sweep it under the carpet... They knew much
earlier but didn't tell...
Donald Dalley
Ronald van Eijck (ron...@noblehouse.xs4all.nl) wrote:
: In article <<Du1wqG.Gyn...@torfree.net>> Donald Dalley writes:
: > If 64 bit
: >chips/OS were used, the Amiga would definately have an edge in timing and
: >technology. This may be just the ticket.
: >I'd like to see some discussion re: a complete 64-bit Amiga.
: >Donald
: And what might that 64-bit Amiga be then?
: Sofar there are very limited areas that need capabilities beyond 32 bit
: and the addition of some 64bit math routines to the OS would suffice
: for most of them.
This is why I wanted to see some discussion.
: If we are talking about 64 bit (or better) graphics buswidth that has
: nothing todo with a 64 bit Amiga.
I figure sw/hw go together. The wider the bus, the wider the bandwidth,
the more bits, the faster everything goes.
Thanks,
Donald
--
+ Are you experienced? Observant?
+ Send your subway/bus rider tips to ab...@torfree.net .
John Sheehy
hfri...@fix.uni-trier.de (Hans-Joerg Frieden) writes:
>Almost all processors have bugs. The 386 had a problem with div's when
>the number was on a segment boundary. The 68000 had a mask fault that
>made some instructions perform in user mode whereasall further
>generations mad them supervisor mode instructions. There's mailing lists
>with processor bug announcements. The point is the last part of the
>sentence: Intel tried to sweep it under the carpet... They knew much
>earlier but didn't tell...
Perhaps, but I thought the point was how well Motorola tested the chips
before making them final.
<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <jsh...@ix.netcom.com>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
Kolbjørn Barmen
David Corn (dc...@paradise.pplnet.com) wrote:
> I also remember reading this. Now, key question - is this MSs or
> Intel's fault? In other words, do you see this same problem with
> (from 8 -> 9) processors on an Intel box under an Intel Unix?
No, thats the amusing part of it :)
> >Remember unix doesnt neccesserly imply user-unfriendlyness,
> >Thats all up to the developers.
>
> Oh great - and you really think you'll sell with an attitude like
> that?
unixes sell pretty well, you know the network we're communicating through
right now, it's not based on M$ products. No, there are unixes all the way
through, and the sysops along the backbone are not interested in changing
system, so it will keep on being unix. The fact that there are some NT www
and ftp sites spread along doesnt mean much to the backbone.
Even in the microsoft.com domain there are lots of unix machines,
f.ex phoenix.microsoft.com which is a FreeBSD machine... ;)
Daniel Reichen
In article <6609...@jojo.IN-Berlin.DE> "Bernhard Graf" <gr...@jojo.IN-Berlin.DE> writes:
>
> The data bus is 16 bit.
> The registers are 32 bit and the command set offers the appropriate
> 32bit instructions.
>
> Afaik the M68k's ALU is only 16bit. Therefore 32bit operations take more
> cycles than with 16bit ops.
>
In the A3000 the architecture is 32 bit straight through memory, Zorro3,
everything. This is why it came with the '030, a 32 bit CPU.
The '040 and '060 both are 32 bit also.
The '000 and '020 are 16 bit.
TTYL
Daniel
------------------------------------------------------------------------
BERN Enterprises http://www.aracnet.com/~bern
Micro Music Publishing INet: be...@aracnet.com
Amiga Computer Tutor
Voice: (503) 940-4789
Northwest Amiga Group Home Page http://www.aracnet.com/~bern/NAG
Matthias Andree
In 31DEDA...@tribeca.ios.com, je...@tribeca.ios.com (Jean Klos) wrote:
> The 680x0 all have 16 bit instruction words; hence the "16/32 bit"
> designation.
>
> As far as all of the other posts are concerned, some ask why Motorola
> still sells older 680x0, whereas Intel doesn't sell many 80x86. Simple.
> How functional is an old Amiga or an old Mac? Not very functional, but
> not completely dead. Most software will still run on old 68000s. Try
> running anything on an 8086 or a 286. Intel blew up many bridges and
> still have a crappy 8 bit instruction core in every Pentium made.
Don't think so. Why should anyone who has his CPU only gather data get a
Pentium Pro 200 if a 68020 with 16 MHz is appropriate?
--
MATTHIAS ANDREE Amiga 4000 EC030@25/882@33 14M 1G5
studying electrical engineering at the University of Dortmund
PGP public key available via mail or via FINGER man...@sx1.hrz.uni-dortmund.de
Jonathan Belson
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >: Atari is 64bit. That's not debatable (well, it shouldn't be ;-)
: >:
: >
: >What I had heard is that the Atari Jaguar used two 32-bit CPUs, and that
: >really isn't a *true* 64-bit machine.
:
: But it has 2 64 bit graphic coprocessors and a 64 bit memory address space.
Edge magazine gave a detailed breakdown of the machine when it came out. I
don't have the mag to hand but I'm quite sure that it had two 32-bit
co-processors, which Atari's marketing division decided to call 64-bit.
Explains its rather lacklustre performance, doesn't it? 8^)
C-YA
Jon
bme...@bruce.cs.monash.edu.au
gn...@morgan.ucs.mun.ca (George Noel) writes:
>In article <4rn695$c...@coranto.ucs.mun.ca>,
>Byron Montgomerie <bmon...@morgan.ucs.mun.ca> wrote:
>>So the amiga should have a 64 bit (data bus) motherboard at what clock
>>rating? You willing to pay for that? If it were cost effective the PC
>If DEC doesn't have a cost reduced version of their Alphas, then at least
>300 Mhz. :)
Just as a side note --- you can get 166MHz ALPHA systems with a full
64 bit OS preinstalled for $1199. It's a tough market.
bme...@bruce.cs.monash.edu.au
gn...@morgan.ucs.mun.ca (George Noel) writes:
[the Atari Jaguar]
>But it has 2 64 bit graphic coprocessors and a 64 bit memory address space.
Boy, I sure am glad that the Jaguar can address more than those pityful
4 Gigabyte.....
bme...@bruce.cs.monash.edu.au
hfri...@fix.uni-trier.de (Hans-Joerg Frieden) writes:
>Michael Kraemer (kra...@clri6j.gsi.de) wrote:
>: They don't need to. At the rate they are pumping out new x86 generations
>: it's not worthwhile to get a replacement. If your '486 breaks, dump it
>: and get a '586 instead. Motorola in contrast needs/needed much more time to
>This would mean that the 586 would easily plug into a 486 socket, but it
>doesn't. And as far as I know, Intel does not make _any_ 586 at all,
>these are Cyrix chips...
Oh, they do. It's just that the Intel 586 isn't a processor, but rather
an some obsolete ethernet controller or some such, named at a time when
nobody ever thought about taking the 80x86 series to x=5. I have actually
seen one of them not too long ago, on a junkyard. It was almost worth
keeping....
David Corn
On 7 Jul 1996 23:07:27 GMT, ko...@nvg.unit.no (Kolbjørn Barmen) wrote:
>David Corn (dc...@paradise.pplnet.com) wrote:
>
>> I also remember reading this. Now, key question - is this MSs or
>> Intel's fault? In other words, do you see this same problem with
>> (from 8 -> 9) processors on an Intel box under an Intel Unix?
>
>No, thats the amusing part of it :)
I see. It was my understanding that you did see this same problem.
>unixes sell pretty well, you know the network we're communicating through
By what measurement do Unix boxes/OSs sell "pretty well"?
>right now, it's not based on M$ products. No, there are unixes all the way
>through, and the sysops along the backbone are not interested in changing
>system, so it will keep on being unix. The fact that there are some NT www
>and ftp sites spread along doesnt mean much to the backbone.
>Even in the microsoft.com domain there are lots of unix machines,
>f.ex phoenix.microsoft.com which is a FreeBSD machine... ;)
I'm familiar with the backbone of the Internet. The machines type
hasn't been changed in years; there's nothing new there. But for
every Unix backbone, there are thousands (hundreds of thousands?) of
PCs and Macs, and a few dozen other computers. :)
Unix is not a common OS.
David Corn
On 7 Jul 1996 19:13:50 GMT, mur...@nmia.com (Trepanation) wrote:
>Lightwave; all I hear about NT is that, once you get it configured
>(assuming you haven't set off a thermonuclear device out of frustration),
>it is indeed powerful.
Frustration? NT is dead simple to set up. Why do you think it's
frustrating? Have you set up WinNT on your machine?
David Corn
On 7 Jul 1996 19:27:13 GMT, hfri...@fix.uni-trier.de (Hans-Joerg
Frieden) wrote:
>Michael Kraemer (kra...@clri6j.gsi.de) wrote:
>This would mean that the 586 would easily plug into a 486 socket, but it
>doesn't. And as far as I know, Intel does not make _any_ 586 at all,
>these are Cyrix chips...
I'd thought the AMD 586/133 (nee 486) plugged into the 486 socket. Is
that not the case? Hmmm.....
George Noel
In article <31df73b2....@news.onramp.net>,
David Corn <dc...@paradise.pplnet.com> wrote:
>On 7 Jul 1996 06:28:56 GMT, gn...@morgan.ucs.mun.ca (George Noel)
>wrote:
>
>>It is starting to unfold this year.. heck even Microsoft put off putting
>>a Windows '96 out due to lack of expected Windows '95 sales.
>
>hands within the next few months. It's been in beta and user review
>for months now.
What they were going to call Windows '96 was released afterwards as a fix
pack only. The last I heard is that they are working on Windows '97. I
forget my source of information but I will try and find it again.
>>>Who care what Intel does, if they drop the ball there then Cyrix or AMD will
>>>take up the slack.
>>
>>But their CPUs are crap compared to Intel and even Intel is not the best
>>CPU makers out there.
>
>Really? I think Cyrix does a very good job with CPUs - their 150+
>models are $260 at local shops, with a good motherboard and 256k
Hadn't you heard that their Pentium clones (earlier ones anyway) ran
slower then the Pentium they said they were equal to and that their CPUs
never had an inbuilt math co-pro etc.
>>No, I expect the whole industry is going to slow down now due to
>>saturation and poor sales. It has started in the memory production area
>>and I doubt Intel are going to want to piss off their buyers again this
>>soon by putting out another optimised version of their CPU this soon.
>
>:) You obviously haven't heard about the MMX then? :)
Actually I have. I meant new ones after the current ones they are
working on. If they release the MMX series and the P7's so soon, it will
be interesting to see what the sales will be like.
-=*George*=-
George Noel
In article <31DF73...@concentric.net>,
Brian Turmelle <bri...@concentric.net> wrote:
>George Noel wrote:
>>
>> I see that as the only logical step to take unless you go with
>> multiprocessing as the other course. PCs time have come.. with the such
>> a large downfall in buying of PCs this year due to the large system
>> resources you need to upgrade to just to run Windows or the latest
>> a lead into 64 is too little, too late? PCs are still using a CISC
>> once the HP/Intel RISC chip comes out, Intel do not guarantee that old
>> programs will run under it. This means PC users either start aknew with
>> the new processor or check out a different system.. either way, most of
>> their old software will not work.
>
World wide, yes. Large system resources such as in some cases upgrading
motherboards, CPUs, hard drives and memory just to run Windows '95 let
alone games through it.
>still selling very very strong. My fathers 486/33 with 8MB RAM didn't need any
Perhaps in North America but even say Compaq (I think it was) felt the
slowdown, Apple took losses (I know, not PC clone but still a computer
company) and look at Escom in Europe!
>upgrading to run Windows 95. I'll probably upgrade it to a P133+ with 16MB RAM
Yeah and how well does it run under that spec? Virutal memory would have
to be turned on in his 486 (slowing it down even more) and how big of a hard
drive does he have?
>for Christmas but that won't be very expensive at all($300). My friends 486/66
Does the motherboard have a 32 bit or 64 bit data bus?
>that's 3 years old runs Windows 95 just fine and most of the newer games. BTW,
How many megs of memory? Virtual memory turned on or off?
>there are PC's now that use a RISC design(Cyrix's Pentium clones). P7's are
>already in silicone, and will surely be out before AmigaOS gets ported. If
>Cyrix can make a RISC Intel clone that's 100% compatable I'm sure Intel can
Is it a fully RISC design though?
>too. Intel has bent over backwards to remain fully compatable. I highly doubt
>most of it will not run.
Yeah, the p7's might be out before AmigaOS gets ported but as for running
software, from what I have heard (forget the source right now) is that
there will be a lot of compatibility problems until new versions of the
programs are written.
>And as we all know, it doesn't matter how well designed a system or processor
>is. It depends on the software. Also, megahertz alone doesn't mean a thing.
Yes software support is very important and all Amiga related companies
should aim at winning developers back to their platforms.
>that runs 25% faster at the same Mhz. PPC has a good design but it isn't fast
>enough to compete with Intel. They don't have the resources, even with IBM,
>Motorolla and Apple combining together. There's a company that's got some
It has been shown that a PPC processor at the same clock speed as an
Intel processor runs faster then Intels.
>new pattents for BiCMOS fabricating that's supposed to double current PPC
>speeds, but I'm sure Intel can get into that too. Intel is adding more things
But the haven't yet so that other company (whose name I forget also) will
beat them to it.
>to keep the Pentium line ahead. Like the 86 MMX (Multimedia Extensions) to the
>X86 instruction set.
But is having everything done via the processor smart? For the company it
is I guess but for the computer overall has yet to be seen. They might
start hurting graphic accelerator card companies then.
-=*George*=-
George Noel
In article <31df686d....@news.onramp.net>,
David Corn <dc...@paradise.pplnet.com> wrote:
>On 7 Jul 1996 04:44:02 GMT, gn...@morgan.ucs.mun.ca (George Noel)
>wrote:
>>>The 060 is the low end. It's where the Amiga should start.
>>
>>Yeah right.. have you seen the price for the processor alone? PPC chips
>>are cheaper then the 060.
>
>Great - then go PPC. So who's going to write this new OS, and do you
>think it will be finished by the year 2000?
There is already plans to go PPC by PIOS and PhaseV but for VIScorp,
there is perhaps one or two more years left with an 68030 or 68040 Amiga
computer and years beyond that for their set-top box equivalent.
As to whose going to write the OS, so far it seems it will be a
co-development between PIOS and VIScorp (I am not sure how PhaseV's work
in progress will fit into that) and it is said to take one year (by late
1997) to get it ported and updated somewhat.
-=*George*=-
George Noel
In article <4roil7$2...@due.unit.no>, Kolbjørn Barmen <ko...@nvg.unit.no> wrote:
>Brian Turmelle (bri...@concentric.net) wrote:
>
>> MMX (Multimedia Extensions) to the X86 instruction set.
>
Yes, Commodore's collaboration with HP would have been nice!
>than half clockspeed, and also the forthcomming PA-9000, which is meant to
>be a Intel/HP hybrid, with full i386 compatibility.
^^^^^^^^^^^^^^^^^^
That's where you get "Not all older programs running".
>--
> Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
> ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
-=*George*=-
George Noel
In article <4roqeq$5...@coranto.ucs.mun.ca>,
Byron Montgomerie <bmon...@morgan.ucs.mun.ca> wrote:
>George Noel (gn...@morgan.ucs.mun.ca) wrote:
>: >Unless you have VISCorp stock, who cares about their set top box? So you
>
>: If they plan on selling millions of them like they say they are going to,
>: then that will bring back developers to the Amiga/ED.
>
>ED developers, not amiga developers, there is a difference.
Not really.. same chips, same OS (all improvements will be used in the
Amiga computer as well) except the ED will have fastram as standard until
VIScorp up the specs on the A1200. As long as they do not take full
advantage of the extra aspects of the ED or makeup for the parts in the
A1200 that are not in the ED, then some simple changes in the code will
make it work on both platforms.
>: >think an 030 or 040 (expensive) aga+ amiga is the appropriate next step in
>: >the amiga line? If they sell them at $300 or so perhaps, which would be
>
>: For VIScorp, within the next 1 year? Yeah. Yes, for a low price point
>: also.. anywhere from $300 - $500 US or so.
>
>Won't happen.
We'll have to see, perhaps not.
>PCs won't be around forever no, but only through evolution into something
>more advanced. There are those that say the dinosaurs didn't all just keel
>over and die, they evolved into other creatures like birds. :)
Good example as the underdog mammals at the time quickly took over the
world! :) IBM's monopoly fell, AT&T's monopoly fell, Intel and
Microsoft's monopoly will fall.
>: It is starting to unfold this year.. heck even Microsoft put off putting
>: a Windows '96 out due to lack of expected Windows '95 sales.
>
>Yeah right, and people will stop using computers suddenly too. :)
Computers as we know it.. that is what IBM, Apple, SUN, Netscape and
Oracle want... they want 50% of the market from Intel and Microsoft by
the year 2000 via their new vision of a PC and Network computer.
>: >: Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
>
>: >For this month.
>
>: No, I expect the whole industry is going to slow down now due to
>: saturation and poor sales. It has started in the memory production area
>: and I doubt Intel are going to want to piss off their buyers again this
>: soon by putting out another optimised version of their CPU this soon.
>
>You'll be disappointed then, they have to keep improving or they will lose
>market share to Cyrix.
In my opinion, they all will be losing market share! :)
>Regards,
>
>BM
-=*George*=-
Brian Turmelle
George Noel wrote:
>
> >: I see that as the only logical step to take unless you go with
> >: multiprocessing as the other course. PCs time have come.. with the such
> >: a large downfall in buying of PCs this year due to the large system
> >: resources you need to upgrade to just to run Windows or the latest
> >: games. They are only now getting into 32 bit so how can you say getting
> >
>
> It is starting to unfold this year.. heck even Microsoft put off putting
> a Windows '96 out due to lack of expected Windows '95 sales.
Where did you get the idea Microsoft was planning on putting out a Windows 95
successor in 96? According to Microsoft's own site, they didn't plan on putting
out a significant upgrade until 97. The only new operating system due out
relatively soon is NT 4.0. Windows 95 sold faster than any other OS ever.
> >: design, by next year Amiga compatibles will be using a RISC design and
> >: once the HP/Intel RISC chip comes out, Intel do not guarantee that old
> >: programs will run under it. This means PC users either start aknew with
> >: the new processor or check out a different system.. either way, most of
> >: their old software will not work.
> >
> >Who care what Intel does, if they drop the ball there then Cyrix or AMD will
> >take up the slack.
>
> But their CPUs are crap compared to Intel and even Intel is not the best
> CPU makers out there.
Yah right. Cyrix now offers an excellent competitor for the Pentium line. All the
way up to 200Mhz. AMD is further behind but they both offer processors that are 2
or more times faster than a 68060, and at a fraction of the cost. Whether Intel
is the best or not can be debated forever, but they are the most profitable and
have the most sales by far.
> >: Top of the line PC today is around 200 Mhz with the Pentium Pro. DEC
> >
> >For this month.
>
> No, I expect the whole industry is going to slow down now due to
> saturation and poor sales. It has started in the memory production area
> and I doubt Intel are going to want to piss off their buyers again this
> soon by putting out another optimised version of their CPU this soon.
Poor sales? Just think how many Intel machines are in the world. What do you
think those people are going to upgrade to? Faster Intel machines! Intel has to
keep putting out faster processors, because the performance gap between the Intel
clones is getting smaller. Intel makes it's money by getting it out 1/2 a year
ahead of the competition and making a nice premium on it. When a clone begins to
match it they drop their premium to make it competitive.
The Pentium started at 60Mhz and has been optimized to 200Mhz. The Pentium Pro
started at 150Mhz and has already reached 200Mhz. It will definately reach
300Mhz, probably before the end of 97.
Brian Turmelle
Byron Montgomerie wrote:
>
> So the amiga should have a 64 bit (data bus) motherboard at what clock
> rating? You willing to pay for that? If it were cost effective the PC
Um, they do. I believe it was Hewlett Packard that utilized a 64bit data bus to
squeeze some extra performance out of a Pentium Pro. I read it from one of the
magazines on zdnet. Probably Computer Shopper.
> In the last few months the Pentium has gone from 133Mhz as the top speed to
> 200Mhz+, you willing to bet that 64 bit would be enough to compete in 2-3
> years time? Will the dec alpha even be a supported cpu then?
You mean supported by AmigaOS(shrug)? Surely the DEC Alpha will be around. They
have the fastest single CPU and it is being improved as well.
Brian Turmelle
George Noel wrote:
> >
> >So the amiga should have a 64 bit (data bus) motherboard at what clock
> >rating? You willing to pay for that? If it were cost effective the PC
>
> lowend. Like I said, they could be doing it to offer their customers
> servers to run online services off of, with the servers running AmigaOS.
> Eventually a long way down the road, people could have this type of
> computer as the main computer in their home, with seperate lowend Amigas
> in practically each room networked with the main server. :) Yes, I am
> dreaming now. :)
He asked what frequency the bus on the motherboard would support at 64 bits
wide. The CPU is 300Mhz but the data bus on the motherboard definately isn't.
> >In the last few months the Pentium has gone from 133Mhz as the top speed to
> >200Mhz+, you willing to bet that 64 bit would be enough to compete in 2-3
> >years time? Will the dec alpha even be a supported cpu then?
>
> Compete - yes. And the quickness in upgrading their CPUs has hurt
> Intel. As for DEC being supported then.. if they don't do something like
> this soon, perhaps not.
The DEC Alpha has been faster than Intel from the beginning. That still hasn't
hurt Intel. The DEC Alpha is aimed at an entirely different market. Mainly
massively parallel supercomputers (CRAY) and very highend workstations. Intel
has actually started breaching into Alpha territory with the Pentium Pro. If
anything, Intel is hurting Alpha, not the other way around.
Brian Turmelle
George Noel wrote:
> True and I don't think Intel are going to be pumping out too many
> versions of their CPU for the next while. The jump from the 60, 75, 90,
> 133, 166 and 200 MHz CPUs within 14 months really caused a lot of people
> to sit back and say WHOAA! Slow down! And companies to write off massive
> amounts of hardware making them lose money when no one was buying the
> newer CPUs. No wonder PCs are so cheap.
>
The increase in speed is due to newer .35 micron fabs. The design hasn't been
changed since the fix for the FDIV bug. It still has the same number of
transistors. The P60 started at .65 micron. And the first runs at the Pentium
Pro were done at .65 micron. Designing a chip on paper and implementing it in
silicon is a big step. It takes time for the fabrication problems to be
perfected. There are .25 and smaller fabs in the works. The whole P5 line is
made with the same process. The amount of imperfections in the transistors is
what determines whether it's 200Mhz or 60Mhz. They are thoroughly tested at
200Mhz first. If it passes it's stamped 200Mhz. If it doesn't it's tested at
166Mhz. If it passes it's stamped 166Mhz. etc etc. If there's demand for
P100's, and most of the chips tested 133+ Mhz, leaving no 100Mhz. You'll get
a P133 for the price of a P100, because the cost to Intel is the same for
the most part.
George Noel
In article <Du7I7...@wombat.hanse.de>,
<bme...@bruce.cs.monash.edu.au> wrote:
>gn...@morgan.ucs.mun.ca (George Noel) writes:
>
>[the Atari Jaguar]
>
>>But it has 2 64 bit graphic coprocessors and a 64 bit memory address space.
>
>Boy, I sure am glad that the Jaguar can address more than those pityful
>4 Gigabyte.....
Would ya believe a 64 bit data bus! :)
>Bernie
>--
>==============================================================================
>Still thinking the Scots should have gone through to the quarterfinals.....
-=*George*=-
George Noel
In article <19960707.72...@BernEnt.com>,
Daniel Reichen <be...@aracnet.com> wrote:
>In article <6609...@jojo.IN-Berlin.DE> "Bernhard Graf" <gr...@jojo.IN-Berlin.DE> writes:
>
>>
>> The data bus is 16 bit.
>> The registers are 32 bit and the command set offers the appropriate
>> 32bit instructions.
>>
>> Afaik the M68k's ALU is only 16bit. Therefore 32bit operations take more
>> cycles than with 16bit ops.
>>
>
>In the A3000 the architecture is 32 bit straight through memory, Zorro3,
>everything. This is why it came with the '030, a 32 bit CPU.
>
>The '040 and '060 both are 32 bit also.
>
>The '000 and '020 are 16 bit.
Are you sure the '020 is 16 bit? I thought it was 32 bit. If it is 16
bit why was the CD32 touted as the "World's first 32 bit game system"
when it had pretty much the makings of a stripped down A1200 inside?
>TTYL
>
> Daniel
>
>------------------------------------------------------------------------
>BERN Enterprises http://www.aracnet.com/~bern
>
>Micro Music Publishing INet: be...@aracnet.com
>Amiga Computer Tutor
> Voice: (503) 940-4789
>
> Northwest Amiga Group Home Page http://www.aracnet.com/~bern/NAG
-=*George*=-
Brian Turmelle
> > can do? If you did then you wouldn't squibble over 12MB RAM. NT is not
> > designed for a single user. It is a multiuser, preemtive multitasking
> > multithreading powerhouse. It even supports multiple processors. We
> > have a Compaq here at work that has 4 P133's in it, with 256MB RAM.
>
> And how many users in at once?
It acts as a fileserver, database server and app server. Over 200 on an
average day.
> > Windows NT runs on x86, DEC Alpha and PowerPC processors.
>
> *blurp*
>
> Linux/i386
> Linux/alpha
> Linux/m68k (amiga, atari, mvme)
> mkLinux/PPC for Apple Macintosh
> Linux/PPC
> Linux/SPARC
> Linux/MIPS
> Linux/ARM (acorn)
> Linux/SMP
Which supports how many professional applications that a corporation might
use?
> NetBSD-alpha
> NetBSD-m68k (mac, atari. amiga. mvme, sharp)
> NetBSD-arm32 (acorn)
> NetBSD-hp300 (HP 9000/300 series)
> NetBSD-i386 (family PC)
> NetBSD-pc532 (PC532)
> NetBSD-pica (Acer Pica)
> NetBSD-pmax (DECstation)
> NetBSD-sparc (Sun SPARC series)
> NetBSD-sun3 (Sun 3)
> NetBSD-vax (DEC VAX)
>
> And they're for free!!
Neither of which support Microsoft Office, Microsoft Backoffice, Oracle, or
any other professional tool of that quality.
> "Gå hjem og vogg" as we say it in Norway.
>
> > Of course it wasn't fully optimized for each processor.
> > If AmigaOS gets anywhere near as powerful as NT then we'll see who has
> > a bloated OS.
>
> "Gosh.. we're real impressed down here"
Apparently you are because you've dumped the AmigaOS argument and jumped to
Unix variants that have been compiled for the 68k series.
> Apropos multiple processors.. NT shows a flattening curve in performance
> when you add more CPUs, after 8 CPUs the system spends just as much
> more CPU-time to to administrate the signalprosessing as the the extra
> CPU is capable of doing. _Real_ multiple OSes show a stright line in
> performance when adding more CPUs, even after CPU number 8.
_Real_ such as? And at what cost? I've taken advanced operating systems
courses. Having 2 processors never equals 2X performance. The performance
increase always follows a curve. The point is that it does support multiple
processors, even if 4 is the max because of cost vs effectiveness. If each
processor adds 66% on average, that's 3 times the performance of a single
CPU system. Three times the power of a fast Intel CPU which is already
several times faster than the fastest CPU AmigaOS will run on.
> Source? Dont remember, some OpenWorld some months ago or so I think...
Doesn't really matter. When it comes to multiprocessor OS's, the hardware
has a significant role in how added processors increase performance. Who
knows what system they tested it on.
> Whats this posting's got to do with the amiga? Well, for one the above
> unixes, Linux and NetBSD, are running on amigas per today, and they're
> getting better from day to day. I recommend people with HD-space, MMU and
> some amount of patients to check them out. At least it's quite educative.
No idea. Some Amiga advocate probably made yet another false claim about NT,
and it evolved into this.
I like unix in general. Linux runs very well on my P133.
> Second, it's the above OSes AmigaOS people should look at when creating
> the core of the "new" AmigaOS, not some bloated unix-wanna-be OS from some
> dodgy softewarehouse infameous for it's buggy software, and aggressiv
> marketing techniques.
I don't think anyone suggested that the new AmigaOS should be like NT.
A bloated unix-wanna-be OS? Call it whatever you wish, I'll take it over
AmigaOS any day. It's comments like those that get PC advocates to come back
and trample on what little life the Amiga has left in it. Suit yourself.
> Make AmigaOS a fully loaded bsd for PPC (and perhaps even DEC Alpha), and
> I would have no doubt on what platform I will choose in the future.
> Remember unix doesnt neccesserly imply user-unfriendlyness,
> Thats all up to the developers.
If you like Unix so much, and you like the the horsepower of a DEC Alpha,
you can get a DEC Alpha box and load Solaris or Linux or whatever flavor of
Unix you like best. And better yet, you can do it today.
Brian Turmelle
Kolbjørn Barmen wrote:
>
> Sounds like inchip hardware 386-emulation, tsk tsk..
It's a lot more sofisticated than that. It's a RISC core that can handle the
Intel instruction set. Superpipelined, out of order execution, branch
prediction, multiple pipelines, multiple ALUs, etc. Check http://www.cyrix.com
for complete details.
> > It's how many instructions it can process in a CPU cycle that matters.
>
> MIPS? Also known as BOGOMIPS? ;)
I didn't say how many of the quickest instructions it can process in a CPU
cycle. =P
> > The only real serious alternative is a company like Cyrix (IBM
> > fabricates them, competing with themselves) that has a pin compatable
> > Pentium(not Pro) processor that runs 25% faster at the same Mhz. PPC
> > has a good design but it isn't fast enough to compete with Intel. They
> > don't have the resources, even with IBM, Motorolla and Apple combining
> > together. There's a company that's got some new pattents for BiCMOS
> > fabricating that's supposed to double current PPC speeds, but I'm sure
> > Intel can get into that too.
>
> I have a theory they've already done so. The fastes pentium today do not
> handle much overclocking before saying bye bye, not that this really
> indicates anything..
The speed increases in the Pentium series from 60Mhz to 200Mhz is due to going
from .65 micron to .35 micron fabs. As well as the fabs gaining experience
over the years to get out most of the imperfections in the fabrication
process.
> > Intel is adding more things to keep the Pentium line ahead. Like the 86
> > MMX (Multimedia Extensions) to the X86 instruction set.
>
> Hm.. also there is HP's PA-8000 which outraces the DEC Alpha with less
> than half clockspeed, and also the forthcomming PA-9000, which is meant to
> be a Intel/HP hybrid, with full i386 compatibility.
Don't forget the MIPS R10000.
> > Of course processors will get faster as production facilities perfect
> > their operations. The original pentium started at 60Mhz and has now
> > tweaked to 200Mhz. The Pentium Pro started at 150Mhz and is also
> > supposed to achieve 300Mhz-400Mhz with die size reductions and
> > "tweaking". I forget what the P7 is supposed to start at. I'm curious
> > what Cyrix will come out with next.
>
> And the HP-PA8000@180MHz outruns a DEC-Alpha@400MHz, times are turbulent,
> anything can happen.
Neither of which will be a viable alernative until they drop into the price
range of the Intel series. Which I don't see happening any time soon. The PPC
seems to have the best overall chance right now.
Brian Turmelle
> > just to lower the memory requirements for a program by half a megabyte
> > in not the path to the future. It might be cheaper to just supply that
> > memory expansion with the software :-)
>
> Take a look at netscape, they use this technique, adding more and more
> features, without doing any optimalization.
> Notice the tendency of warp from v1.12 up to 3.x-Gold of Navigator?
So you expect Netscape to jump into the field of writing compilers too? You
make optimalization sould like some magical word that will make a program
with 10 times more features shrink down to something smaller than the
original. I'm sure they've optimized whatever they reasonably could.
Michael M. Rye
On Sun, 07 Jul 1996 21:02:31 GMT jsh...@ix.netcom.com (John Sheehy) wrote:
:hfri...@fix.uni-trier.de (Hans-Joerg Frieden) writes:
[snip]
HF:>with processor bug announcements. The point is the last part of the
HF:>sentence: Intel tried to sweep it under the carpet... They knew much
HF:>earlier but didn't tell...
JS>Perhaps, but I thought the point was how well Motorola tested the chips
JS>before making them final.
--
Gee, you could say the same thing about the Intel Pentium fdiv (or whatever
it was) bug.
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
* -Michael M. Rye- (Jedi) | Amiga 1200, 68030/MMU 50 MHz, 68882 50 MHz, *
* UNIX/C Admin/Design |____ 10 MB RAM, 2.0 Gig HD, SupraFAX 28.8 Kbps ____*
* <mr...@topcity.mn.org> | Amiga 500, AdRAM 540 - 1 Chip/2 Fast, 2 floppies *
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
* Thought of the Day: *
* I'm your father, Luke! Give in to the Dark Side, you knob!!! *
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
Michael M. Rye
On Sun, 7 Jul 96 18:33:40 be...@aracnet.com (Daniel Reichen) wrote:
>In the A3000 the architecture is 32 bit straight through memory, Zorro3,
>everything. This is why it came with the '030, a 32 bit CPU.
>The '040 and '060 both are 32 bit also.
True.
>The '000 and '020 are 16 bit.
Huh? Where are you pulling this out of??? As far as I know, the 68000 is
32-bit internally, with a 24-bit address bus and 16-bit data bus. So the
chip itself is 32-bit, while everything outside it is less. Can't remember
about the '020. (Someone correct me if I'm wrong.) Check out Motorola's Web
site for more info.
>TTYL
> Daniel--
Michael Kraemer
In article <4rk9i3$p...@due.unit.no>, ko...@nvg.unit.no (Kolbjørn Barmen) writes:
> Michael Kraemer (kra...@clri6j.gsi.de) wrote:
> >
> > They don't need to. At the rate they are pumping out new x86 generations
> > it's not worthwhile to get a replacement. If your '486 breaks, dump it
> > and get a '586 instead. Motorola in contrast needs/needed much more time to
> > and 68060. You still can buy 68020/30/40 chips and probably also some
> > derivative of the 68000. Don't know if intel still makes 8086/80186/80286/80386,
> > probably not.
>
> And what is the marked for intels?
n*100 millions of PCs of course. For most people CPU == x86, now that the 68k
has lost the desktop (and the workstations too, for that matter).
> 68k's you can find in old VAXes, fridges, TVs, Filmcameras, Communication
?????
> equipment, jetfighters, spaceshuttles, sattelites, soundstudios, scanners
> broadcaststudios, lightcontrollboards, cars, printers, monitors, etc...
Yes, but nobody cares about it. And embedded controlling slowly but steadily
is moving to PowerPC, away from 68k. I've even seen ads for industrial
controllers that are x86 driven.
> They make quality prosessors that need warranty, they also run heavy tests
> on them before releasing them on the marked.
Yes, OK, but on the other hand this attitude led to the loss of the desktop market.
AFAIK the first IBM PC could have been driven by the 68000, if Motorola could
have guaranteed in-time delivery in masses.
The same story for the 68040. It came far too late. Most workstation makers
went from the 68030 to some RISC design because they needed power and the 68040
wasn't there when it was needed.
And again the 68060. One of the last 68k followers, Apple, went PowerPC
because the 68060 again was too little too late. Especially too late.
Now the few Amigas are the only desktop computers using the 68060.
> We all remember the fdiv-bug in the first intel pentiums, and how Intel
> tried to sweep it under the carpet ;)
No, nobody remembers. They have the monopoly and they can do what they want to.
They have some prime directive to double the x86 performance
and halve their price every 18 months. They have to, because the x86 line is
their cash business.
> --
> Kolbjørn Barmen /// "I met the Amiga and fell in love ..."
> ko...@nvg.unit.no \XX/ http://www.nvg.unit.no/~kolla/
Not to get me wrong, as an occasional assembly programmer I like the 68k,
it's more fun than any other CPU I've seen so far, but this doesn't count
that much these days. Computing nowadays is a business, not a hobby as it
was in 80's.
Michael Kraemer
In article <Du7Ky...@wombat.hanse.de>, bme...@bruce.cs.monash.edu.au writes:
> hfri...@fix.uni-trier.de (Hans-Joerg Frieden) writes:
>
> >Michael Kraemer (kra...@clri6j.gsi.de) wrote:
> >: They don't need to. At the rate they are pumping out new x86 generations
> >: it's not worthwhile to get a replacement. If your '486 breaks, dump it
> >: and get a '586 instead. Motorola in contrast needs/needed much more time to
> >doesn't. And as far as I know, Intel does not make _any_ 586 at all,
> >these are Cyrix chips...
>
> an some obsolete ethernet controller or some such, named at a time when
> nobody ever thought about taking the 80x86 series to x=5. I have actually
> seen one of them not too long ago, on a junkyard. It was almost worth
> keeping....
>
Ah, so this was the *real* reason that they named it "Pentium" instead of "80586"?
Michael Kraemer
In article <4rp32h$r...@news01.uni-trier.de>, hfri...@fix.uni-trier.de (Hans-Joerg Frieden) writes:
> Michael Kraemer (kra...@clri6j.gsi.de) wrote:
> : They don't need to. At the rate they are pumping out new x86 generations
> : it's not worthwhile to get a replacement. If your '486 breaks, dump it
> : and get a '586 instead. Motorola in contrast needs/needed much more time to
> This would mean that the 586 would easily plug into a 486 socket, but it
> doesn't.
Well, what I meant is if one owns a 486 computer, and this breaks or needs
a substantial upgrade it's not worthwhile to replace it or do the upgrade.
It's cheaper to dump it and buy a new one.
This is radically different from the Amiga market where people keep their
old machines as long as possible. This is why you still can get add-ons
for A500s and software still runs on OS 1.2.
> And as far as I know, Intel does not make _any_ 586 at all,
> these are Cyrix chips...
Well, they call it "Pentium", I guess, but the name it deserves is 586, since
it's just another x86.
> Regards, Hans-Joerg.
>
> --
> Hans-Joerg Frieden |"Why is it, that the years seem to fly past, while
> Schloss-Str. 176 |the night seems to last forever?" - Elminster
> 54293 Trier |"It's like ten thousand spoons when all you need
> hfri...@fix.uni-trier.de | is a knife" - Alanis Morissette
>
Robert Owen Raine
In article <4rr3g4$i...@rs18.hrz.th-darmstadt.de>,
kra...@clri6f.gsi.de (Michael Kraemer) wrote:
>In article <Du7Ky...@wombat.hanse.de>, bme...@bruce.cs.monash.edu.au
writes:
>>
>> >Michael Kraemer (kra...@clri6j.gsi.de) wrote:
>> >: They don't need to. At the rate they are pumping out new x86 generations
>> >: it's not worthwhile to get a replacement. If your '486 breaks, dump it
>> >: and get a '586 instead. Motorola in contrast needs/needed much more time
to
>> >This would mean that the 586 would easily plug into a 486 socket, but it
>> >these are Cyrix chips...
>>
>> an some obsolete ethernet controller or some such, named at a time when
>> nobody ever thought about taking the 80x86 series to x=5. I have actually
>> seen one of them not too long ago, on a junkyard. It was almost worth
>> keeping....
>>
>
>Ah, so this was the *real* reason that they named it "Pentium" instead of
"80586"?
Don't think so the reason that everyone knows about is the fact that Intel
lost there patent case against Cyrix and others. They tried to patent the
numbers 286 386 486 and said that it was copyright infringement. Court said "I
don't think so, you can't copyright numbers." Intel said O.K. we call this new
chip "PENTIUM". Penti greek for 5, get it. Intel said now we can sue there
buts. As usual Cyrix not as stupid as Intel seems to think, called there
Pentium replacement 586 and figured that the rest of us weren't stupid either.
\|/
@ @
------------------------------------------------------oOO-(_)-OOo--------------
Only Amiga Makes it Happen Bob Raine
The computer for the creative mind Michigan State University
Make Up Your Own Mind Physics Astronomy Dept.
Amiga / The Alternative Ra...@msupa.pa.msu.edu
Escom: "Anything is Possible
"Viscorp:Our intent to purchase the Amiga comes out of
(i) an appreciation of the core technology and its importance to VIScorps
long-term business development and
(ii) a belief that the Amiga can -- like the Phoenix -- rise from the ashes
and become a profitable technology for the future.
-----------------------------------------------------------------------------------------
Jean-Michel Bezeau
Bonjour George Noel, sur Re: Motorola and the m68k (was "Re: VIScorp's possible plans..") (message du 08-Juil-96 02:30:02),
GN> In article <19960707.72...@BernEnt.com>,
GN> Daniel Reichen <be...@aracnet.com> wrote:
>>In article <6609...@jojo.IN-Berlin.DE> "Bernhard Graf"
>><gr...@jojo.IN-Berlin.DE> writes:
>>
>>>
>>> The data bus is 16 bit.
>>> The registers are 32 bit and the command set offers the appropriate
>>> 32bit instructions.
>>>
>>> Afaik the M68k's ALU is only 16bit. Therefore 32bit operations take more
>>> cycles than with 16bit ops.
>>>
>>
>>In the A3000 the architecture is 32 bit straight through memory, Zorro3,
>>everything. This is why it came with the '030, a 32 bit CPU.
>>
>>The '040 and '060 both are 32 bit also.
>>
>>The '000 and '020 are 16 bit.
GN> Are you sure the '020 is 16 bit? I thought it was 32 bit. If it is 16
GN> bit why was the CD32 touted as the "World's first 32 bit game system"
GN> when it had pretty much the makings of a stripped down A1200 inside?
020 are 32 bit. From a certain point of vue, 000 are as well: you can
ignore the fact that they are 32/16 and program all in 32 bit, the
Amiga OS is programmed that way, so are most Amiga program since that
was in the developer's guideline from the beginnings.
Returning to the 020, it is a 32 bit CPU. The EC020 is another story,
it is a 32/16 as the 68000 however it doesn't support 16 bit or 24 bit
addressing as the 68000 does.
--
star...@Clic.Net Jean-Michel Bezeau
Le retour des sombres est proche Programmeur/Analyste/Traducteur
Monstre sacré de Baarsk Spécialiste d'Amiga OS
Dragon originel Téléphone: 418-843-1738
Damné éternel Fax: 418-843-3655
--
Jim Wilde
In article <4rgrr8$2...@coranto.ucs.mun.ca>
bmon...@morgan.ucs.mun.ca (Byron Montgomerie) writes:
>George Noel (gn...@morgan.ucs.mun.ca) wrote:
>
>: If they could get this 64 bit technology (don't forget DEC Alpha's START
>: at 64 bit) at a low cost, then imagine the set-top boxes they could make
>: then! :) If Nintendo can put out a lowend 64 bit box for $250 US or so, I am
>: sure VIScorp could do amazing things with that kind of technology in a
>: set-top box and a computer!
>
>Personally, I would rather they went with parrallel cpus before using 64 bit
>cpus. Then again it depends on what you call 64 bit too.
>
>Regards,
>
>BM
>
Can someone tell me why this interminable discussion is taking place in the
multimedia newsgroup?
Byron Montgomerie
George Noel (gn...@morgan.ucs.mun.ca) wrote:
: In article <4roqeq$5...@coranto.ucs.mun.ca>,
: Byron Montgomerie <bmon...@morgan.ucs.mun.ca> wrote:
: >George Noel (gn...@morgan.ucs.mun.ca) wrote:
: >: >Unless you have VISCorp stock, who cares about their set top box? So you
: >
: >: If they plan on selling millions of them like they say they are going to,
: >: then that will bring back developers to the Amiga/ED.
: >
: >ED developers, not amiga developers, there is a difference.
: Not really.. same chips, same OS (all improvements will be used in the
Well, if the amiga is going to be essentialy an ed box then I might as
well take my ass in hand and move on.
: Amiga computer as well) except the ED will have fastram as standard until
: VIScorp up the specs on the A1200. As long as they do not take full
: advantage of the extra aspects of the ED or makeup for the parts in the
: A1200 that are not in the ED, then some simple changes in the code will
: make it work on both platforms.
See you next year in that case when all this repeats itself. :)
: >Won't happen.
: We'll have to see, perhaps not.
Silly rabbit, tricks are for hookers. :)
: >PCs won't be around forever no, but only through evolution into something
: >more advanced. There are those that say the dinosaurs didn't all just keel
: >over and die, they evolved into other creatures like birds. :)
: Good example as the underdog mammals at the time quickly took over the
: world! :) IBM's monopoly fell, AT&T's monopoly fell, Intel and
: Microsoft's monopoly will fall.
You are forgetting that a lot of the other dinosaurs did die out due to
over specialization. :) And that all this took place over millions of
years. :)
: >Yeah right, and people will stop using computers suddenly too. :)
: Computers as we know it.. that is what IBM, Apple, SUN, Netscape and
: Oracle want... they want 50% of the market from Intel and Microsoft by
: the year 2000 via their new vision of a PC and Network computer.
If wishes were fishes...
: >You'll be disappointed then, they have to keep improving or they will lose
: >market share to Cyrix.
: In my opinion, they all will be losing market share! :)
To what? All those hundreds and thousands of companies dependent upon the
PC will just switch directions like a school of fish? :)
Regards,
BM
(Stay tuned for more of Mother Goose's fables from the crypt)
Byron Montgomerie
Jim Wilde (JWI...@ukcc.uky.edu) wrote:
: In article <4rgrr8$2...@coranto.ucs.mun.ca>
Didn't bother to trim the list of newsgroups that some idiot added to this,
sorry for my part in it.
Regards,
BM