Help: AMD k6-3 450, Windows 95 and Asus p5a-b
Moonbase Commander
been there done it before....
Win98 was a lot easier to install then Win95A,B,C so with Win98 want to
install Win98SE but know you have to fine tune after install......
--
Moonbase Commander
http://home.moonbase-alpha.net/p5a.htm
Jim <jimm...@hotmail.com> wrote in message
news:7pqi0c$pst$1...@imsp009a.netvigator.com...
> I'am using K6-3 450 too...but running Windows 98 SE with Voodoo 2000 and
> 2940W SCSI card, everything work fine.
>
> First check all your mother board jumper setting, upgrade board bios to
1007
> and change to windows 98 or 98 SE version
>
Fozia Zaidi
I decided to upgrade my Pentium 233 MMx on a Red Fox Motherboard
to a Asus P5A-b board with AMD k6-3 450 last week and I've been having
countless problems since!
I installed the board and AMD chip then on boot up to windows 95, I
would an
error message: "While initialzing IOS: Windows Protection error" please
reboot
your computer. When I turned the cpu cache off everything appeared to
work fine.
I'm in the process of getting my AMD chip checked out to see if its ok.
I read somethign about a microsoft patch for AMD k6-2 350's. Would that
solve
my problem on the k6-3 450 too?
ALi chipset problems:
basically if I install an of the Ali busmaster drivers than
all my CDrom's dissappear. If I leave them to the generic controller
drivers than
everything "seems" to be ok
But will my performance be lost if I don't install the drivers?
I also tried installing some of the other Asus add on software but for a
lot of them
I get this message: " You need Windows OSR version 2.1 or higher"
my windows version from the My computer properties is 4.00.950B - what
OSR version is that?
I have an ATI all in wonder pro card 128 AGP - haven't seen any
compatibility problems yet.
Anyhow this whole experience has been quite a pain- this is also the
first time
I've installed a motherboard too which might explain some of this.
So can anyone help me out with any configuration errors I might be
making?
Thanks
Fozia Zaidi
Jim
Kelly Lu
don't actually "own" an AMD, but I would suggest trying that patch on the
Microsoft website. I've heard countless problems (too many problems to say
the truth) with AMD and Windows 95.
Hope you get this solved! And next time, stick with Intel.
Kelly Lu
--
Find Hidden Potential in Your Computer!
The Exchange! - http://theexchange.cjb.net
ICQ - 8674004
Fozia Zaidi <f...@geocities.com> wrote in message
news:37C0D2DC...@geocities.com...
Martin Kou || cybermartin
Microsoft's AMD CPU patch, otherwise the the OSR2.1 upgrade will overwrite
the patched files.
First of all, since u can hardly boot into Windows 95 successfully in any
clock speed higher than 350MHz, u'll need to set the clock speed back to
somethin' lower, say, 66MHz x 4.5 = 300MHz, and boot into Windows 95. Go
onto the Internet to search for usbsupp.exe(for upgrade Windows 95 from OSR2
to 2.1), be sure that the language of the file matches the language of your
Windows 95, u can't install an English one to a Chinese Windows.
Reboot and return back the updated Windows 95, get 'n install the file
ftp://ftp.microsoft.com/softlib/mslfiles/amdk6upd.exe, this time the file is
not language sensitive, u should can install it to any language version of
Windows 95 OSR2.1. Shut down the machine, turn the clock back to 100MHz x
4.5 = 450MHz and boot.
This time u should can boot into Windows successfully and now its time to
install the softwares given by Asus.
for the Bus Mastering Drivers problem: have u rebooted after the Windows 95
detected the Bus Master Controllers? If yes, u need to disable DMA access
for the CD-ROM's IDE channel, which should can be done by tweaking the
registry or in CMOS setup. If no, reboot and see what happens.
Martin Kou || cybermartin
<cyber...@hotmail.com>
P.S. sure, the performance will be degraded if u use only the generic
drivers.
Jeffrey Karp
Kelly Lu wrote:
> This is why I don't get AMD chips. Too many problems. Because of this, I
> don't actually "own" an AMD, but I would suggest trying that patch on the
> Microsoft website. I've heard countless problems (too many problems to say
> the truth) with AMD and Windows 95.
> Hope you get this solved! And next time, stick with Intel.
LOL! You should have said that there are too many problems with Windows,
and that next time he should get Linux, as the problem was with Windows 95,
and not any AMD CPUs.
>
> Kelly Lu
>
> --
> Find Hidden Potential in Your Computer!
> The Exchange! - http://theexchange.cjb.net
> ICQ - 8674004
>
Yeah Right
K6-2 333. I had a bad CPU. Since then no problems at all and the system is
very stable. In fact I built a second. As for your CD-ROM problems, try it
and see. Check you system properties performance tab and see if you your
getting 32 bit file systems access. if you are don't worry about it. But
always keep looking for information.
Good luck
John
Fozia Zaidi wrote in message <37C0D2DC...@geocities.com>...
Roger J.Hamlett
>
> Hi,
>
> I decided to upgrade my Pentium 233 MMx on a Red Fox Motherboard
> to a Asus P5A-b board with AMD k6-3 450 last week and I've been
having
> countless problems since!
>
> I installed the board and AMD chip then on boot up to windows 95, I
> would an
> error message: "While initialzing IOS: Windows Protection error"
please
> reboot
> your computer. When I turned the cpu cache off everything appeared
to
> work fine.
> I'm in the process of getting my AMD chip checked out to see if its
ok.
> I read somethign about a microsoft patch for AMD k6-2 350's. Would
that
> solve
> my problem on the k6-3 450 too?
The problem applies to Windows 95, and affects potentially all AMD
chips faster than 333MHz. It also affects the Intel Xeon
processors in their fastest variants. The 'patch' is for W95 OSR2,
and W98 has the fix allready 'built in'. The problem is that MS wrote
a 'timing loop' for a small delay in part of the IDE code, and the
faster chips execute this so quickly, that insufficient delay is
generated. The AMD's are so quick at this integer loop, relative
to the Intel chips, that they hit the problem at lower clock
rates than Intel. Some IDE hardware seems a little 'less prone' to
the problem than others, and occasionally 350MHz chips will be 'OK',
but otherwise the patch should be installed. By turning the cache
off, you slow the processor sufficiently to prevent the problem.
Best Wishes
--
--------------------------------------------------------------------
| EMail ro...@ttelmah.demon.co.uk http://www.ttelmah.demon.co.uk/ |
| A beard! A beard! cried Fly Nicholas.'By God, that's a good one!'|
| (Chaucer) |
Dan Burke
case you didn't know, has trouble with AMD processors over 350 (they're just
too damn fast), and will freeze up. His cpu isn't bad. His motherboard
isn't bad.
I have a Via motherboard, and they come with the "special" hdd controller
drivers, but I don't install them. I've run HDtach before and after
installing them, and even when I had them configurered correctly, the
standard windoes 98 drivers ran faster. While this is under 98, i would
imagine 95 would be similiar, unless your drivers don't support UDMA.
That version, if memory serves me correct, is only OSR 2. If you have a win
95 cd that says something to the effect of having usb support, then you have
2.1. But otherwise, it seems to me that the Asus software is using usb, the
reason it needs the newer os. Of course, you could just go and get the 98
upgrade, or get 2000 when it comes out.
Dan Burke
d...@locovaca.dyn.ez-ip.net
Yeah Right <sa...@nowhere.com> wrote in message
news:7pqvhq$76b$1...@news.kadena.attmil.ne.jp...
> Sounds exactly like a problem I had building a system with the P5A-B and a
> K6-2 333. I had a bad CPU. Since then no problems at all and the system
is
> very stable. In fact I built a second. As for your CD-ROM problems, try
it
> and see. Check you system properties performance tab and see if you your
> getting 32 bit file systems access. if you are don't worry about it. But
> always keep looking for information.
>
> Good luck
> John
> Fozia Zaidi wrote in message <37C0D2DC...@geocities.com>...
> >Hi,
> >
> >I decided to upgrade my Pentium 233 MMx on a Red Fox Motherboard
> >to a Asus P5A-b board with AMD k6-3 450 last week and I've been having
> >countless problems since!
> >
> >I installed the board and AMD chip then on boot up to windows 95, I
> >would an
> >error message: "While initialzing IOS: Windows Protection error" please
> >reboot
> >your computer. When I turned the cpu cache off everything appeared to
> >work fine.
> >I'm in the process of getting my AMD chip checked out to see if its ok.
> >I read somethign about a microsoft patch for AMD k6-2 350's. Would that
> >solve
> >my problem on the k6-3 450 too?
> >
Rich Perna
news:9_4w3.138367$U42.3...@news1.rdc1.bc.home.com...
> This is why I don't get AMD chips. Too many problems. Because of this, I
> don't actually "own" an AMD, but I would suggest trying that patch on the
> Microsoft website. I've heard countless problems (too many problems to say
> the truth) with AMD and Windows 95.
> Hope you get this solved! And next time, stick with Intel.
No thanks (yes, his problem is fixed with the Microsoft patch). I've been
buying Asus P5A-B motherboards with K6-2/350s and 450s and those machines
*never* crash, which is more than I can say for my piece of crap PII boxes.
Those things go down more often then... oh, wait, this isn't a poly-sci
NG... ;-)
Rich
Extreme
"INTEL cpu" sucks. When i can to amd and upgrade to windows 98, it works
perfectly. Long live AMD!
Regards,
\!/
(@) (@)
(_)
----------oOO---------OOo----------
Kelly Lu <kel...@home.com> wrote in message
news:9_4w3.138367$U42.3...@news1.rdc1.bc.home.com...
> This is why I don't get AMD chips. Too many problems. Because of this, I
> don't actually "own" an AMD, but I would suggest trying that patch on the
> Microsoft website. I've heard countless problems (too many problems to say
> the truth) with AMD and Windows 95.
> Hope you get this solved! And next time, stick with Intel.
> Kelly Lu
>
> --
> Find Hidden Potential in Your Computer!
> The Exchange! - http://theexchange.cjb.net
> ICQ - 8674004
>
> Fozia Zaidi <f...@geocities.com> wrote in message
> news:37C0D2DC...@geocities.com...
Roger J.Hamlett
>
> This is why I don't get AMD chips. Too many problems. Because of
this, I
> don't actually "own" an AMD, but I would suggest trying that patch
on the
> Microsoft website. I've heard countless problems (too many problems
to say
> the truth) with AMD and Windows 95.
> Hope you get this solved! And next time, stick with Intel.
> Kelly Lu
run W95 on the faster Xeon's...
BUBBA !2000!
<gary_...@rocketmail.com> wrote:
>I've previously own an intel cpu. It got problem with windows 95. Therefore
>"INTEL cpu" sucks. When i can to amd and upgrade to windows 98, it works
>perfectly. Long live AMD!
>
>
You need to learn how to build a computer.
HTH!
Thomas Veach
>
> Kelly Lu wrote:
>
> > This is why I don't get AMD chips. Too many problems. Because of this, I
> > don't actually "own" an AMD, but I would suggest trying that patch on the
> > Microsoft website. I've heard countless problems (too many problems to say
> > the truth) with AMD and Windows 95.
> > Hope you get this solved! And next time, stick with Intel.
>
> and that next time he should get Linux, as the problem was with Windows 95,
> and not any AMD CPUs.
>
> >
Ah Beng
>I've previously own an intel cpu. It got problem with windows 95. Therefore
>"INTEL cpu" sucks. When i can to amd and upgrade to windows 98, it works
>perfectly. Long live AMD!
>
I have no problem with "Intel No Inside" for years .'.
Kelly Lu
problem? Even though AMD has faster processors out, Intel still has the
"standard" processors that all (or most) programs are designed for. AMD has
to try to immitate Intel with their processors so as they will work with
current software/hardware. The "Windows problem" is one example of the
inaccuracy of immitating other products.
Kelly Lu
--
Find Hidden Potential in Your Computer!
The Exchange! - http://theexchange.cjb.net
ICQ - 8674004
Thomas Veach <tmv...@mmm.com> wrote in message news:37C2A3...@mmm.com...
Jorge F. Delgado
>
> If it's a Windows' problem, then why doesn't Intel's CPU's have the same
> problem? Even though AMD has faster processors out, Intel still has the
> "standard" processors that all (or most) programs are designed for. AMD has
> to try to immitate Intel with their processors so as they will work with
> current software/hardware. The "Windows problem" is one example of the
> inaccuracy of immitating other products.
> Kelly Lu
>
I believe you are ill informed.
It is a Microsoft problem. To put it simple, They had a bug, which
prevented
correct execution on processors executing a certain piece of code in
less
than a specified time. In that period, only Intel processors were to
blame, so
they introduced a code to fix it (in the final product) based on the
Intel's
Id of the processor.
Unluckily they didn't foresee that AMD or any other would come with
processors
that fast, so the _SAME_ behaviour exhibited in beta test by Intel
fastest
processors started appearing in AMD ones, thus Microsoft issued the
patch for
AMD processors.
Jeffrey Karp
Kelly Lu wrote:
> If it's a Windows' problem, then why doesn't Intel's CPU's have the same
> problem? Even though AMD has faster processors out, Intel still has the
> "standard" processors that all (or most) programs are designed for. AMD has
> to try to immitate Intel with their processors so as they will work with
> current software/hardware. The "Windows problem" is one example of the
> inaccuracy of immitating other products.
> Kelly Lu
LOL! There have never been any problems with AMD processors and Linux.
Microsoft used a shortcut that was logically flawed when designing Windows 95.
Intel chips will probably have problems with Windows 95 once they get over
700mhz. Imo, the problem that Windows 95 had with AMD CPUs is quite analogous
to the Y2K problems that many programs have. They simply did not anticipate
future developments, and accomodate for them. Or perhaps they did, but
intentionally left this potential problem there as a nice way of designing
planned
obsolescence into a product. .What did programmers have in mind when
they designed programs vulnerable to Y2K problems? Perhaps intentional
planned obsolescence was the goal.
Roger J.Hamlett
>
> If it's a Windows' problem, then why doesn't Intel's CPU's have the
same
> problem? Even though AMD has faster processors out, Intel still has
the
> "standard" processors that all (or most) programs are designed for.
AMD has
> to try to immitate Intel with their processors so as they will work
with
> current software/hardware. The "Windows problem" is one example of
the
> inaccuracy of immitating other products.
The problem is that AMD's chip is faster on integer arithmetic, than
the equivalent clock rate Intel units. With Intel chips, only the
very fastest Xeon processors display the problem, and since these are
usually supplied for multi-processor OS's which don't have the fault,
the problem is practically unheard of. I met the problem when it
first appeared, and subsequently could display the same fault on a
Xeon 500. I suspect the 600MHz P-3 will also give the fault, but
there is a significant difference in useage patterns. People using
the AMD are often upgradeing, and have the older releases of W95.
With something like the Xeon, this is generally not used, and with
the P-3, most people are buying new machines, or because of a
motherboard change, having to re-install, and are therefore using
W98, which is fixed in this regard. The fault is actually an example
of ludicrously 'bad' programming, using a small timing loop for a
delay that is shorter than conveniently available from the internal
timer 'tick' (not itself unusual), and then failing to take
variations in the processor speed into account (silly). Also though
you say that 'Intel' is the standard, it is worth realising that
parts of DX6, are now written specifically to take advantage of the
3DNow capabilities of the AMD units...
Thomas Veach
>
> If it's a Windows' problem, then why doesn't Intel's CPU's have the same
> problem? Even though AMD has faster processors out, Intel still has the
> "standard" processors that all (or most) programs are designed for. AMD has
> to try to immitate Intel with their processors so as they will work with
> current software/hardware. The "Windows problem" is one example of the
> inaccuracy of immitating other products.
It's well known that Winblows was written around the architecture of the
Intel processor and therefore operates well with Intel. Linux, Unix, and
Even DOS on non-Intel chips, run circles around Winblows 9X on any chip
other than an Intel.
REMOVE...@yahoo.com
wrote:
Yes but since Intel invested money in the biggest Linux distrubutor(
among others...) Red Hat the same thing would be to Linux(Red Hat
version).
Athlon does not run ok on Linux Red Hat if you dont update the kernel!
Anyone heard that service pack5 does stop any overcloking gain when
running latest pentium III :-)
bye
Magnus
Magnus
Eric van Bezooijen
Kelly Lu <kel...@home.com> wrote:
>If it's a Windows' problem, then why doesn't Intel's CPU's have the same
>problem? Even though AMD has faster processors out, Intel still has the
>"standard" processors that all (or most) programs are designed for. AMD has
>to try to immitate Intel with their processors so as they will work with
>current software/hardware. The "Windows problem" is one example of the
>inaccuracy of immitating other products.
>Kelly Lu
Sigh. It has been stated many times in many of these newsgroups that the
reason the AMD chips expose this Windows bug is because they do a certain
loop operation much faster than an Intel chip at the same Mhz. Intel chips
display the same behavior at a higher Mhz. You don't you want AMD to slow
their chips down just because Microsoft writes buggy software, do you ?
-Eric
[ lossy compression ]
>> > > >
>> Roger that! The AMD chip is fine, It's Winblows that has the problem.
--
Eric van Bezooijen er...@activesw.com http://www.activesw.com/~eric
"But the meaning of life is a mystery, that we don't understand so far.
And the music of life is a rhapsody if you're happy the way that you are."
-"Freudiana"<P><IMG SRC="http://www.activesw.com/~eric/images/eric.gif">
Bob May
and if the chip isn't on the list of accepted processors then it is
rejected and the AMD cpu violates that request. I can't believe that
a simple little patch will fix anything but something like that.
Considering the bloat in the program, I dare anybody to prove that the
above isn't happening. It's also interesting that user programs don't
seem to have the problem with the CPU chip!!! It's only the WIN
program itself. If that's not true, why is there no patches for the
various programs like Office and so forth. I would expect that
programs like Autocad and so forth to really have problems if windows
is having a problem.
--
Bob May
I don't read attachments to posts as they may give me a
virus If I expect an attachment from you I will open it..
You may have a brilliant thought but if you put it into an
attachment I won't read it and thus both you and I lose.
I don't like to say it but unfortunatly, there are those who
insist upon being nasty to the rest of us. Bob May
Dan Burke
I think you need to rethink your theory.
Dan Burke
d...@locovaca.dyn.ez-ip.net
Bob May <bob...@access1.net> wrote in message
news:37c2...@news.access1.net...
Guy Southern
Jeffrey Karp <jeff...@erols.com> wrote in message
news:37C2BF53...@erols.com...
>
>
> LOL! There have never been any problems with AMD processors and Linux.
> Microsoft used a shortcut that was logically flawed when designing Windows
95.
> Intel chips will probably have problems with Windows 95 once they get over
> 700mhz. Imo, the problem that Windows 95 had with AMD CPUs is quite
analogous
> to the Y2K problems that many programs have. They simply did not
anticipate
> future developments, and accomodate for them. Or perhaps they did, but
Exactly. It's called "Get it out the door." If you recall, Windows 95 was
late. I suspect the fix for only Intel processors involves the wonderful
"don't break anything else."
> intentionally left this potential problem there as a nice way of designing
> planned
> obsolescence into a product. .What did programmers have in mind when
> they designed programs vulnerable to Y2K problems? Perhaps intentional
> planned obsolescence was the goal.
Don't kid yourself. Y2K has nothing to do with planned obsolescence. Most
of the problems are with Cobol, and were written when space was a premium,
and hardware often didn't support full dates. Even if programmers (I am
one) don't expect their programs to last, they do expect data and code to.
Y2K is shortsightedness.
Microsoft does have planned obsolescence, of course. Look at the way they
handled USB in Win95 and NT4.0...
Guy
RedHat
REDHAT
Doug Rose
98
http://www.walbeehm.com/win95upd.html#UpdList
BTW-I'm still using osr2.0
This page has a short list For Ver A's and a Short list for version B's and
C's.
Go into safe mode-check for 'ghost devices' -extra CD-rom Devices in Device
Manager
and make sure no config.sys of autoexec.bat device drivers are loading for
cd-rom.
Make sure you install ALI bridges and patches-on the Asus CD
or download the latest.
One of them is the AGP patch -must install MS "USBSUPP.EXE"
and another is the ALI powermanagement -1.5 OR 1.6 (has been a while)
-even if you have to disable the MS-powermanagement.
Also do the above updates and fixes on the OS FOR YOUR VERSION.
Cannot take Win 950a to Win950b. Period.
Yes, you won't get the Ultra DMA speed without them.
However, could try Mitsumi CD-rom Drivers-they work for just about
any CD-Rom. Or use the setup disk that came with your Win98-if that is your
OS.
-generic oak cd-rom drivers there.
1. BusMaster Drivers-no dos drivers for CD-rom.
2. Power Management ALi Patch-M7101 or something like that.
3. USBsupp.exe-from microsoft or other pages- or ftp search engine
3. AGP ALi patch.
OSR2.1= Windows 95 Ver B.-system file dates =August 1996
I think 08-24-96-then "USBSUPP.EXE"
The "USBSUP.EXE"- is the update to the first and don't know if
it has everything.
Lastly, in bios enable USB (under "PNP" section), assign a free IRQ
or "auto". Then system will recognize your OS as 2.1. The ".1" is
USBSUPP.EXE
installed.
Doug Rose
>The problem is that AMD's chip is faster on integer arithmetic, than
>the equivalent clock rate Intel units. With Intel chips, only the
>very fastest Xeon processors display the problem, and since these are
>usually supplied for multi-processor OS's which don't have the fault,
>the problem is practically unheard of. I met the problem when it
>first appeared, and subsequently could display the same fault on a
>Xeon 500. I suspect the 600MHz P-3 will also give the fault, but
>there is a significant difference in useage patterns. People using
>the AMD are often upgradeing, and have the older releases of W95.
>With something like the Xeon, this is generally not used, and with
>the P-3, most people are buying new machines, or because of a
>motherboard change, having to re-install, and are therefore using
>W98, which is fixed in this regard. The fault is actually an example
>of ludicrously 'bad' programming, using a small timing loop for a
>delay that is shorter than conveniently available from the internal
>timer 'tick' (not itself unusual), and then failing to take
>variations in the processor speed into account (silly). Also though
>you say that 'Intel' is the standard, it is worth realising that
>parts of DX6, are now written specifically to take advantage of the
>3DNow capabilities of the AMD units...
>
>Best Wishes
>| A beard! A beard! cried Fly Nicholas.'By God, that's a good one!'|
>| (Chaucer) |
>
-Doug
Jeffrey Karp
Kelly Lu wrote:
> Ok, ok... some of these posts in this group are very convincing. And I am
> convinced that it's a Windows bug, not an AMD problem. But, I am still going
> to stick with Intel CPUs because Intel is a much larger company with ample
> resources to test and make sure their processors are in top shape when they
> ship.
LOLOLOLOL!!!! ROFLMAO!!!!!!!!
That is the funniest thing I have read all week. AMD never shipped any chips
that
that performed calculations with math errors! You seem to have it backwards.
Intel has pushed processors out the door as rapidly as possible whether they
calculated
with math errors or not, while AMD not only verifies its own chips, but
thoroughly
tests motherboards made by others that accept AMD chips. That is why Athlon
systems
were delayed by a few weeks. AMD was not happy with the initial motherboards
made
for the Athlon. New improved motherboards have now been released, with AMDs
approval for use with Athlons.
> AMD, correct me if I'm wrong, is a much smaller company, and because
> of a high demand for their "low cost alternatives" they don't spend as much
> money or effort to ensure their processors are high quality products.
LOL!!! That is so funny! Intel has done a great job of brainwashing you.
If Intel really spent that much verifying their chips, do you think they
would have released many that performed calculations with math errors?
Did you ever consider why Intel spent so much on the Intel inside ad
campaign? Intel was smart enough to know that AMD would eventually
come out with CPUs that were more technologically advanced than their own.
To protect their market share, they felt they needed to brainwash many people.
I can honestly say they did a good job on you.
> This
> could be the reason for a lot of the "bad" AMD cpus out there.
There are not many bad AMD CPUs out there. There are many bad super
7 motherboards, however the problems people have with AMD CPUs
and bad motherboards are usually the result of people not using motherboards
that are on AMDs recommended list.
> Now, Intel
> also has these "bad" cpus, but I have not run into any, whether in first
> person, second person, or third person. Wowever, the AMD "bad" processor, I
> have run into, with a few of my friends, and some people in this newsgroup.
> That, IMHO, is the reason why large corporations depend on Intel processors
> and not AMD.
I can't stop laughing. You comments are so entertaining!
>
> This has turned out to be a great discussion!
Indeed it has!
>
> Kelly Lu
>
> --
> Find Hidden Potential in Your Computer!
> The Exchange! - http://theexchange.cjb.net
> ICQ - 8674004
>
> Fozia Zaidi <f...@geocities.com> wrote in message
> news:37C0D2DC...@geocities.com...
> > Hi,
> >
> > I decided to upgrade my Pentium 233 MMx on a Red Fox Motherboard
> > to a Asus P5A-b board with AMD k6-3 450 last week and I've been having
> > countless problems since!
> >
> > I installed the board and AMD chip then on boot up to windows 95, I
> > would an
> > error message: "While initialzing IOS: Windows Protection error" please
> > reboot
> > your computer. When I turned the cpu cache off everything appeared to
> > work fine.
> > I'm in the process of getting my AMD chip checked out to see if its ok.
> > I read somethign about a microsoft patch for AMD k6-2 350's. Would that
> > solve
> > my problem on the k6-3 450 too?
> >
Kelly Lu
convinced that it's a Windows bug, not an AMD problem. But, I am still going
to stick with Intel CPUs because Intel is a much larger company with ample
resources to test and make sure their processors are in top shape when they
of a high demand for their "low cost alternatives" they don't spend as much
could be the reason for a lot of the "bad" AMD cpus out there. Now, Intel
also has these "bad" cpus, but I have not run into any, whether in first
person, second person, or third person. Wowever, the AMD "bad" processor, I
have run into, with a few of my friends, and some people in this newsgroup.
That, IMHO, is the reason why large corporations depend on Intel processors
and not AMD.
Dan Burke
intel, so they trust it. Just like the schools and apple computers: even
though apple was on the brink of bankruptcey, they still bought apple
computers because they trusted the name and company.
Dan Burke
d...@locovaca.dyn.ez-ip.net
Kelly Lu <kel...@home.com> wrote in message
news:euGw3.144816$U42.3...@news1.rdc1.bc.home.com...
chris pitzel
>to stick with Intel CPUs because Intel is a much larger company with ample
>resources to test and make sure their processors are in top shape when they
Very few problems with AMD-based setups have been attributed to the actual
CPU itself. AMD CPU's on Intel-chipset-based motherboards are just as
reliable as the genuine Intel items.
However, once you stray into the land of Super Socket 7 boards, and the
various 3rd party chipsets (Ali, VIA, SiS, etc.), the problems begin.
>ship. AMD, correct me if I'm wrong, is a much smaller company, and because
>of a high demand for their "low cost alternatives" they don't spend as much
>money or effort to ensure their processors are high quality products. This
The processor, in general, is not the root of the problem in the first
place. It's the damm chipsets. The SS7 chipsets progessively have been
getting better with more recent steppings, but 99% of the problems
reported in the newsgroups can be attributed to a specific chipset, and
fixed by use of a different chipset.
>could be the reason for a lot of the "bad" AMD cpus out there. Now, Intel
>also has these "bad" cpus, but I have not run into any, whether in first
>person, second person, or third person. Wowever, the AMD "bad" processor, I
Ever heard of the floating-point division bug? Or the various errata with
the Pentium Pro/P6 processors? There's no reason to believe that the AMD
errata list is any longer than the Intel errata list on a given part.
>have run into, with a few of my friends, and some people in this newsgroup.
>That, IMHO, is the reason why large corporations depend on Intel processors
>and not AMD.
My belief is that the chipsets that people tend to use with AMD chips
these days are more problematic than the CPU itself could ever be.
Intel has the advantage of having an immensley popular, single chipset
'family' (i440BX) available for their PPro class of processors.
Gary Sanford
wrote:
>Roger that! The AMD chip is fine, It's Winblows that has the problem.
Actually the AMD's are more stable than the Intel's IMO.
~~~~~~~~~~~~~~~~
Gary Sanford
sanf...@ibm.net
Gary Sanford
wrote:
I regard "Intel Inside" like the warning on cigarettes.
~~~~~~~~~~~~~~~~
Gary Sanford
sanf...@ibm.net
Darren Winsper
> Well then, if thats the case, where's the IDT patch?
Show me an IDT running at 350MHz first.
> And the cyrix patch?
Show me a Cyrix running at 350MHz first.
> I think you need to rethink your theory.
I think you need to do the same because you're making yourself look
like a fool.
--
Darren Winsper - http://easyweb.easynet.co.uk/darren.winsper
Stellar Legacy project member:
http://easyweb.easynet.co.uk/darren.winsper/stellar
"God - error 38522: God does not exist - Warning 71154: the following
characters have been ignored: God." - Some mainframe with a crisis of faith.
Stephen Chadfield
> My suspicion is that windows has some code to test for a INTEL chip
> and if the chip isn't on the list of accepted processors then it is
> rejected and the AMD cpu violates that request. I can't believe that
> a simple little patch will fix anything but something like that.
> Considering the bloat in the program, I dare anybody to prove that the
> above isn't happening. It's also interesting that user programs don't
> seem to have the problem with the CPU chip!!! It's only the WIN
> program itself. If that's not true, why is there no patches for the
> various programs like Office and so forth. I would expect that
> programs like Autocad and so forth to really have problems if windows
> is having a problem.
A specific part of the Windows code has problems when running at high
speeds. Why does this mean there should also be bugs in applications?
Windows is software just like the user applications and can have bugs
independantly of those programs.
--
Stephen Chadfield
David
Darren Winsper wrote in message ...
>On Tue, 24 Aug 1999 19:53:02 GMT, Dan Burke <d...@locovaca.dyn.ez-ip.net>
wrote:
>> Well then, if thats the case, where's the IDT patch?
>
>Show me an IDT running at 350MHz first.
>
>> And the cyrix patch?
>
>Show me a Cyrix running at 350MHz first.
>
>> I think you need to rethink your theory.
>
>I think you need to do the same because you're making yourself look
>like a fool.
Firstly I think u shouldn't be calling other people fools when it seems
to me that you may have misunderstood what Dan Burke was saying. Bob May had
said that he(Bob) thought that Windows tests a cpu to determine if it is an
intel or not and if it wasn't an intel Window rejects it and hence the bug.
Dan was merely trying to point out that if that was the case, the bug would
also affect cyrix and idt chips (ie all non-intel chips). That seems to me
to have nothing to do with the speed rating of any of the chips.
David
chrisv
wrote:
>That is the funniest thing I have read all week. AMD never shipped any chips
>that performed calculations with math errors! You seem to have it backwards.
>Intel has pushed processors out the door as rapidly as possible whether they
>calculated with math errors or not
Haven't they? AMD chips have errata, like almost all CPU's, do they
not?
While I agree that the person you were responding to had some pretty
silly justifications for preferring Intel over AMD, I'm not sure that
Intel's Pentium FPU problem of 5 years ago is strong evidence that
Intel makes bad products. The problem with the FPU was somewhat
"understandable," after all, although their initial handling of the
situation was not well done.
Rich Perna
news:euGw3.144816$U42.3...@news1.rdc1.bc.home.com...
> Ok, ok... some of these posts in this group are very convincing. And I am
going
> to stick with Intel CPUs because Intel is a much larger company with ample
> resources to test and make sure their processors are in top shape when
they
Remember the F00F bug? I don't recall AMD being hit with math problems in
their CPUs. Also, a lot of the "compatibility" issues are board
manufacturers (e.g., nVidia) not properly testing on Super7 motherboards.
Granted, this issue wasn't cleared up until recently, but now that board
makers realize how large the Super7/Intel alternative market is, they're not
going to ship product without proper testing first. The TNT fiasco taught
them all a lesson they're not going to forget.
>AMD, correct me if I'm wrong, is a much smaller company, and because
> of a high demand for their "low cost alternatives" they don't spend as
much
> money or effort to ensure their processors are high quality products. This
> could be the reason for a lot of the "bad" AMD cpus out there. Now, Intel
There is *not* a problem with "bad" AMD cpus. There is a problem with
low-quality motherboards, but this affects the Intel market as well. The
mobo is NOT a place to cut costs when putting a system together,
unfortunately a lot of people don't realize it. The most reliable
motherboards IMO are made by Asus - I've never had a problem with any of
their products. For a Taiwan-based company, they have incredible support in
the USA. They're very responsive to their e-mail, and their fast reaction
to the customer outcry regarding Athlon motherboards is proof that they are
a company that listens. I've had problems with various motherboards from
other companies (the only one I can think of I've never had any experience
with is Tyan), especially voltage difficulties. It doesn't matter whether
you're putting an Intel or AMD CPU in your machine, if your motherboard is a
piece of crap you're system is going to be unstable.
> also has these "bad" cpus, but I have not run into any, whether in first
> person, second person, or third person. Wowever, the AMD "bad" processor,
That's because unless CPUs are abused, they're fairly well tested, much more
so than most PC components. All CPU manufacturers MUST burn-in and
thoroughly test their CPUs, otherwise how do they know what speed they're
going to be reliable at for the customer? I have built hundreds of systems,
and have *never* seen a "bad" CPU. I also use a static strap while putting
systems together, this could be why. :-)
> have run into, with a few of my friends, and some people in this
newsgroup.
> That, IMHO, is the reason why large corporations depend on Intel
processors
> and not AMD.
Remember, AMD chips haven't been in "mainstream" PCs for all that long.
Also, the most popular PC manufacturers with corporations are Compaq and
Dell, neither of whom put AMD CPUs in their "business" machines. This will
hopefully change with the Athlon, and I'd put much better odds on AMD making
it into business desktops than Linux. (damn, can of worms, let me screw the
lid back on!)
Rich
Roger J.Hamlett
>
> My suspicion is that windows has some code to test for a INTEL chip
> and if the chip isn't on the list of accepted processors then it is
> rejected and the AMD cpu violates that request. I can't believe
that
> a simple little patch will fix anything but something like that.
> Considering the bloat in the program, I dare anybody to prove that
the
> above isn't happening. It's also interesting that user programs
don't
> seem to have the problem with the CPU chip!!! It's only the WIN
> program itself. If that's not true, why is there no patches for
the
> various programs like Office and so forth. I would expect that
> programs like Autocad and so forth to really have problems if
windows
> is having a problem.
disassembled by many people. There is no 'processor ID' checking in
this area. It is just that the AMD is faster than expected. Latter
Intel chips also display the same problem. It never existed in older
OS's, that used the BIOS for their disk I/O, and does not exist in
latter OS's (W98 WNT, & W2000). It is only W95 that has the fault,
and it is purely a 'bad' bit of programming...
--
--------------------------------------------------------------------
Basil Brush
news:euGw3.144816$U42.3...@news1.rdc1.bc.home.com...
> Ok, ok... some of these posts in this group are very convincing. And I am
> convinced that it's a Windows bug, not an AMD problem. But, I am still
going
> to stick with Intel CPUs because Intel is a much larger company with ample
> resources to test and make sure their processors are in top shape when
they
> ship.
Athlon? These things don't get down on their knees when the clock rate hits
600!
>AMD, correct me if I'm wrong, is a much smaller company, and because
> of a high demand for their "low cost alternatives" they don't spend as
much
> money or effort to ensure their processors are high quality products.
Athlon?
> This could be the reason for a lot of the "bad" AMD cpus out there. Now,
Intel
> also has these "bad" cpus, but I have not run into any, whether in first
> person, second person, or third person. Wowever, the AMD "bad" processor,
I
> have run into, with a few of my friends, and some people in this
newsgroup.
> That, IMHO, is the reason why large corporations depend on Intel
processors
> and not AMD.
Eric van Bezooijen
Rich Perna <ri...@ranir-dcp.com> wrote:
>Kelly Lu <kel...@home.com> wrote in message
>news:euGw3.144816$U42.3...@news1.rdc1.bc.home.com...
>> Ok, ok... some of these posts in this group are very convincing. And I am
>> convinced that it's a Windows bug, not an AMD problem. But, I am still
>going
>> to stick with Intel CPUs because Intel is a much larger company with ample
>> resources to test and make sure their processors are in top shape when
>they
>> ship.
>
>their CPUs. Also, a lot of the "compatibility" issues are board
The AMD K6 had a problem with it (The problem showed up easily on Linux systems
with 64 Mb of memory or more if I recall correctly), and they had a pretty bad
attitude about it just like Intel had with their problems. I got my K6-200
replaced though. Neither company is perfect on this count.
-Eric
Doug Rose
HEY!
Anybody solve or answer shortly the first QUESTION.
Or did we get sidetracked-and the original 'questioner' is on top of
a building with a rifle somewhere near the HQ of Micro-cough.
Short answer-can't be done.
Upgrade to Win98.
SOMETHING-HU!
Doug Rose
>Yes but since Intel invested money in the biggest Linux distrubutor(
>among others...) Red Hat the same thing would be to Linux(Red Hat
>version).
>Athlon does not run ok on Linux Red Hat if you dont update the kernel!
>Anyone heard that service pack5 does stop any overcloking gain when
>running latest pentium III :-)
>
>bye
>Magnus
>
>
>Magnus
Don't know about service pack5, but that might explain why
PIII 450 are less expensive than PII 450.
"The cheapest new thing will beat the best old thing, unless someone has a
monopoly in mind."
Doug Rose
the first K6's came out. Remember how the old Amd 486-40's and so on up to
the K6
ran really hot-and goofy?-Wouldn't touch them then-no pun intended.
Had to do with AMD facilities were not clean enough-particle dust level to
get the micron level.
I don't know-maybe that has changed.
-Doug
Kelly Lu wrote in message ...
>Ok, ok... some of these posts in this group are very convincing. And I am
>convinced that it's a Windows bug, not an AMD problem. But, I am still
going
>to stick with Intel CPUs because Intel is a much larger company with ample
>resources to test and make sure their processors are in top shape when they
>ship. AMD, correct me if I'm wrong, is a much smaller company, and because
>of a high demand for their "low cost alternatives" they don't spend as much
>money or effort to ensure their processors are high quality products. This
>could be the reason for a lot of the "bad" AMD cpus out there. Now, Intel
>also has these "bad" cpus, but I have not run into any, whether in first
>person, second person, or third person. Wowever, the AMD "bad" processor, I
>have run into, with a few of my friends, and some people in this newsgroup.
>That, IMHO, is the reason why large corporations depend on Intel processors
>and not AMD.
>This has turned out to be a great discussion!
>Kelly Lu
>
Anthony Hill
>If it's a Windows' problem, then why doesn't Intel's CPU's have the same
>problem?
The timing loop that causes an error executes essentially
twice as fast on an AMD K6 line of processors then it does on Intel's
P6 core. Since this extremely simple loop should scale almost 100%
linearly with processor speed, it should take a PIII that is twice as
fast (in terms of MHz at least) as an AMD K6-2 to cause the problem.
AMD ran into the problem somewhere between 333 and 350MHz, so Intel is
likely to hit the same problem between 666 and 700MHz, when they reach
that speed. Actually, Intel was going to have hit such a problem a
long time ago with 300 or 333MHz processors, however they contacted
Microsoft and informed them of the problem, and Microsoft "fixed" it
before the initial release of Win95. Unfortunately Microsoft's "fix"
only pushed the speed at which the problem would occur up to about
twice of what it was (my guess is that they probably just did
something simple like changing a signed integer to an unsigned integer
to double the range of a variable).
> Even though AMD has faster processors out, Intel still has the
>"standard" processors that all (or most) programs are designed for. AMD has
>to try to immitate Intel with their processors so as they will work with
>current software/hardware.
AMD has done an EXTREMELY good job at immitating Intel
processors, which is why compatability problems are nearly
non-existant. However, they most certainly can NOT just copy Intel's
design, a VERY large court trial several years back decided this quite
conclusively. As a result, AMD designs their own chips. Different
chip designs obviously do things a little differently. Some things
are done MUCH faster on some designs, while others are done MUCH
slower.
> The "Windows problem" is one example of the
>inaccuracy of immitating other products.
True, but it's also not a problem that is limited to AMD.
Intel has to immitate products as well. Do you truely think that the
Pentium is 100% compatible with the 486? They most certainly aren't!
The differences are small and usually not noticed, but they do exist.
Intel's broken just as many timing loops as AMD has, probably even
more so in their time. I just ran into one a month and a half ago. A
brand new Intel Celeron system absolutely refused to run an old tax
program on my parents computer. My mom is a part time tax accountant,
and in Canada at least, tax records need to be kept for 7 years.
Anyway, to make a long story short, she needed to access records from
1995 or '96, but the tax program for that year died on startup. The
solution? Run a little utility to take up 65% of the systems
resources so that the program ran way slower, then it worked fine.
Anyway, by your reasoning, AMD and Intel should never increase the
speed of their processors because it might break timing loops?
Anthony Hill
hi...@storm.ca
Anthony Hill
wrote:
>Kelly Lu wrote:
>> If it's a Windows' problem, then why doesn't Intel's CPU's have the same
>> "standard" processors that all (or most) programs are designed for. AMD has
>> to try to immitate Intel with their processors so as they will work with
>> inaccuracy of immitating other products.
>
>LOL! There have never been any problems with AMD processors and Linux.
Hehe, nice comment, unfortunately it's not accurate :< The
one true bug that AMD did have with their initial K6 processor (which
was quickly fixed in hardware) mainly just affected Linux systems.
The very first K6 processors had a small bug that only cropped up when
dealing with more then 32MB of memory at a time. Due to Windows, umm,
less then ideal memory architechture, it never really dealt with that
much memory at a given time, however Linux could (and did)
occasionally do so. The only time I ever heard of this happening was
when compiling the kernel in Linux. FWIW though, the problem was not
very widespread (AMD's yields at the time when these processors were
released were piss-poor, so they weren't able to make/sell many), and
AMD did replace any chips for people who were affected by the problem.
Anyway, for those that think that these chips are all error
free, you should check the "errata" sheets for them sometime. These
sheets describe "errors in documentation", which really means that the
chip was supposed to do something one way, but in reality it does
things slightly differently, however the difference is so small that
it's much easier to simply change how it's documented then to change
the processor itself. I haven't looked at AMD's sheets, but the last
time I checked the errata sheet of the Intel Pentium was some 200
pages long.
Anthony Hill
hi...@storm.ca
Darren Winsper
[Me making an idiot out of myself.
> Firstly I think u shouldn't be calling other people fools when it seems
> to me that you may have misunderstood what Dan Burke was saying. Bob May had
> said that he(Bob) thought that Windows tests a cpu to determine if it is an
> intel or not and if it wasn't an intel Window rejects it and hence the bug.
D'oh! For some reason, I thought he was talking about the person
claiming AMD were crap because they made Windows crash.
John P
> Run a little utility to take up 65% of the systems
> resources so that the program ran way slower,
it's called Windows!
Sorry Anthony - couldn't resist!
John
.......
// John Portwin <j...@post.com>
// Cheaply personalise your Nokia Mobile Phone!
// 700+ logos 210+ ringtones http://www.pmbbs.demon.co.uk
Brian Ballard
processors.... the 810 chipset IIRC. And that was very recent.
chrisv <chr...@dgii.com> wrote in article
<37c3e97d...@news.dgii.com>...
Justin Colson
AMD is nolonger a "low cost alternative". Just because AMD is a smaller
company(they admit that they are) doen't mean that their products are worse,
Borland products are much better than Microsoft ones but Microsoft is a much
bigger company.
Kelly Lu <kel...@home.com> wrote in message
news:euGw3.144816$U42.3...@news1.rdc1.bc.home.com...
chris pitzel
I wouldn't say 'much better'. Benchmarks only show a very miniscule
difference between the Athlon and the P3's at identical clock speeds. In
some server-type applications, the P3 is still faster.
>AMD is nolonger a "low cost alternative". Just because AMD is a smaller
>company(they admit that they are) doen't mean that their products are worse,
>Borland products are much better than Microsoft ones but Microsoft is a much
>bigger company.
The Athlon is likely a superb processor. Hopefully the chipsets that are
available for the Athlon don't leave a lot to be desired. AMD's K6
processors have been somewhat hobbled by the low-quality of the Super
Socket 7 chipsets and motherboards.
KALLE
>HEY!
>Anybody solve or answer shortly the first QUESTION.
Answer : Run the AMD patch for windows 95 b!
KALLE
see http://www.jump.net/~lcs/kalle/index.htm for T2P4 and some other information
Jeffrey Karp
chris pitzel wrote:
> >AMD Athlons are much better than Pentium IIIs and even Pentium III Xeons.
>
> I wouldn't say 'much better'. Benchmarks only show a very miniscule
> difference between the Athlon and the P3's at identical clock speeds. In
> some server-type applications, the P3 is still faster
Have you read the reviews of the Athlons?
Here is a link to a pages with links to many in depth reviews.
http://www.jc-news.com/pc/AMD/Athlon_Reviews/
In most applications heavy in floating point, and 3D graphics,
the Athlon is at least a full speed grade faster(ie Athlon 500
is faster than P3 550)
> .
>
> >AMD is nolonger a "low cost alternative". Just because AMD is a smaller
> >company(they admit that they are) doen't mean that their products are worse,
> >Borland products are much better than Microsoft ones but Microsoft is a much
> >bigger company.
>
> The Athlon is likely a superb processor. Hopefully the chipsets that are
> available for the Athlon don't leave a lot to be desired.
They should be great, as they are being made by AMD(later others will
make them as well)
> AMD's K6
> processors have been somewhat hobbled by the low-quality of the Super
> Socket 7 chipsets and motherboards.
That is why AMD has a list of recommended motherboards for their
socket 7 processors.
http://www1.amd.com/K6/k6mbl
http://www.amd.com/products/cpg/tips.html
Martin Kou || cybermartin
<snip>
Martin Kou || cybermartin
While Micro-cough leads the fools, and preachs foolishness.
While AMD is for the wise, ...
p.s. don't u feel Micro-cough hides the deeper information of Windows and ur
system to prevent u from understanding computers and their bugs?
Gary Sanford <sanf...@sprynet.com> wrote in message
news:37c68bc8...@nntp.sprynet.com...
KALLE
wrote:
>i think the lazy programmers are those high-paid in Microsoft.
Not only at Mickysoft.
Turbo Pascal for example was hit long time before. :-)
Sebastian Kaliszewski
> True, but it's also not a problem that is limited to AMD.
>Intel has to immitate products as well. Do you truely think that the
>Pentium is 100% compatible with the 486? They most certainly aren't!
>The differences are small and usually not noticed, but they do exist.
>Intel's broken just as many timing loops as AMD has, probably even
>more so in their time.
That is very known problem of Borland's Turbo Pascsal 7.0 and applications
created using that and any Intel product faster than 200-233MHz. Especially
many shareware programs (as well as not too big commertial apps few years
ago) were made using that compiler. Fortunately there are many generic fixes
available for that. One of the best is simple resident application that
catches the problem on the fly and fixes that. Similar thing happened with
TP 6.0 and 486s! Fortunately bad timing loop didn't crash the programs, but
counted incorrectly causing 1s delay to be only ~0.5s long on i486 33MHz! On
faster CPUs it's even worse. Conclusion: programmesr should never use timing
loops or at least have >100000% safety margin on fastest chip available.
Regards
Sebastian Kaliszewski
--
"Never underestimate the power of human stupidity" - from Notebooks of L.L.
BUBBA !2000!
<mya...@is.mentioned.in.letter> wrote:
>Agreed, high price with low quality, Intel's for fools.
>While Micro-cough leads the fools, and preachs foolishness.
>While AMD is for the wise, ...
>
Please! your ignorance is showing. AMD and Intel both make good
processors, and the celeron is certainly not "high price" You people
should start a religeon (both Intel fanatics and AMD zealots)
KALLE
>MS is noted for NOT paying high wages.
>
>BUT they do give out stock options, which is why so many of those who worked
>long hours (no overtime) became millionaires.
>
>You could look it up!
I answered to lazy programmers, not how they are payed. :-)
KALLE
>My suspicion is that windows has some code to test for a INTEL chip
Replace your suspicion with knowledge:
There is a loop command that programers like to use to build time
loops. Cause there are a lot of lazy programers out there who don't
take care of fast CPU's in their time loops (overflow error) Intel
didn'd optimized this command (or in other words : slowed it down) so
it takes 5 (or was it 6???) cycles to run it.
AMD didn'd care about this lazy programmers and optimized this command
like everything else. So it takes just one cycle to 'operate' it.
That's the reason why WIN95 crashes with AMD K6-350 and Intel
P II-1750.
Lazy Programers!
No room for any conspiracy theory!
Risk Manager
Celeron would not come.
Celeron 300A would not come
100Mhz bus would not come
Intel would control all of the computer user.
The processor world would not be change suddenly.
Moreover, your question is Microsoft and Intel problem.
Not related with AMD.
Thomas Veach wrote:
> Jeffrey Karp wrote:
> >
> > Kelly Lu wrote:
> >
> > > This is why I don't get AMD chips. Too many problems. Because of this, I
> > > don't actually "own" an AMD, but I would suggest trying that patch on the
> > > Microsoft website. I've heard countless problems (too many problems to say
> > > the truth) with AMD and Windows 95.
> > > Hope you get this solved! And next time, stick with Intel.
> >
> > LOL! You should have said that there are too many problems with Windows,
> > and that next time he should get Linux, as the problem was with Windows 95,
> > and not any AMD CPUs.
> Roger that! The AMD chip is fine, It's Winblows that has the problem.
Anthony Hill
wrote:
>Didn't IBM take over the manufacturing of the AMD's around the time when
>the first K6's came out.
AMD signed up IBM as a potential foundry, but IBM never
produced a single K6 chip. They did, however, help AMD get their own
manufacturing process up to par.
> Remember how the old Amd 486-40's and so on up to
>the K6
>ran really hot-and goofy?-Wouldn't touch them then-no pun intended.
>Had to do with AMD facilities were not clean enough-particle dust level to
>get the micron level.
>I don't know-maybe that has changed.
The heat thing wasn't due to any problems with manufacturing
so much as simple design. The K6 core drew quite a bit of current and
the K6 233 ran at a fairly high voltage for a .35 micron chip (3.2V
core, vs. 2.8 or 2.9V core for most other .35 micron chips). AMD's
own fab did have significant productions problems in the early days of
the K6 though. This really prevented them from breaking into the
market until more then 6 months after the K6 was released because AMD
simply couldn't supply enough chips for any major OEM. Later AMD had
significant trouble shrinking their die to .25 micron, and have had a
few smaller problems since then. However since March of this year,
their production has been really top-notch. AMD chips do still run
hot, however as mentioned above, a lot of this is a simple matter of
design. They draw a lot of current (the K7 at 600MHz draws somewhere
around 30A when running full out), and as such need to be cooled quite
a bit.
As for Intel, you don't get to hear about any production
problems from them. They're a large enough company that they can hide
most production problems that do occur, plus manufacturing processors
has always been Intel's strong suit (which is good, because they
aren't all that hot when it comes to designing the chips, especially
considering the resources they've got). Intel recently has had at
least one fairly major problem with manufacturing, they're having a
lot of trouble shrinking their die to .18 microns. They've had to
push the release date of .18 micron chips back by two months. Not
nearly as long as some delays that AMD has experienced, but still not
a particularly good sign for Intel (who's having enough trouble with
all their processor designs being at least 18 months behind schedual).
Anthony Hill
hi...@storm.ca
Martin Kou || cybermartin
effectiveness(i don't know if the phrase is that), AMD CPUs certainly win
that. if u say Athlon is a bit more expansive than PIII, then plz compare
them with Xeon at least, 'cause they have 200MHz FSB and the RAM available
now is just PC100 which means that there is still a large room for
improvement on Athlon's speed.
secondly, if u can feel that, Intel was increasing their speed of advancing
the clock speeds, they certainly have money to research, so what's that for?
the first is of course, to get rid of AMD, Cyrix and the past NeXTgen. the
second, many think, to get rid of your pocket.
so i said they R fooling u...
BUBBA !2000! <RED...@TRAILERPARK.COM> wrote in message
news:37c667fc...@news.supernews.com...
Dan
>> intentionally left this potential problem there as a nice way of
designing
>> planned
>> obsolescence into a product. .What did programmers have in mind when
>> they designed programs vulnerable to Y2K problems? Perhaps intentional
>> planned obsolescence was the goal.
>
>Don't kid yourself. Y2K has nothing to do with planned obsolescence. Most
>of the problems are with Cobol, and were written when space was a premium,
>and hardware often didn't support full dates. Even if programmers (I am
>one) don't expect their programs to last, they do expect data and code to.
>Y2K is shortsightedness.
Erm... That's true... Or at least it is for most code the Y2K
troubleshooters have to deal with.
But there is code written in the eighties (and even some from the
nineties...) that is not Y2K compliant. Christ, go to Microsoft's site and
you'll find (minor) Y2K patches for Win'98!
My point is that it became standard practice to represent dates in the YY
format, even when it was obvious that it would cause severe problems at the
end of '99, and computers had become more than capable of handling 4-digit
years. You have to remember you only need sacrifice one thing (speed or
space) to get 4-digit years. You could either use 4 bytes (for ASCII or
BCD - years 0000 to 9999) or 2 bytes (absolute value - years 0000 to 65535).
Neither of these formats is particularly troublesome to program.
I used to work for a travel software firm that did indeed use Y2K for
obsolescence. They were planning to just dump customers who stuck with
their 'old' system and migrate to client-server. Sod any company that
couldn't afford the 'upgrade'. They were *deliberately* not dealing with
the issue. It is a case of upgrade or lose your system. Stiffing at it's
best, I'd say.
I'd love to know how the company is getting on now most businesses are aware
of the issue... :)
>
>Microsoft does have planned obsolescence, of course. Look at the way they
>handled USB in Win95 and NT4.0...
Agreed. Even the Win'95 OSR2.1 support for USB wasn't perfect by any means.
Regards
Daniel Jones
>
>Guy
>
* * Chasman
involved with a project that was doing reverse engineering on ALL of the
Intel CPUs that were available at the time. I asked if AMD was trying to
save money on development costs and the response was that they were trying
to protect themselves from lawsuits for patent and copyright infringements.
R&D costs AMD a lot more than it does Intel because they have to design
chips that perform the same operations as Intel CPUs but are designed
differently enough so that they don't violate any licensing agreements,
patents or copyrights.
Chas.
Dionysus
<chris....@nospam.usask.ca> wrote:
>>AMD Athlons are much better than Pentium IIIs and even Pentium III Xeons.
>
>I wouldn't say 'much better'. Benchmarks only show a very miniscule
>difference between the Athlon and the P3's at identical clock speeds. In
>some server-type applications, the P3 is still faster.
whereas other benchmarks show a huge difference. so what? it's all
relevant to who you trust....and granted, the athlon isn't much
quicker on integer math, but the FPU (which, for gamers, translated
directly into FPS) is a considerable improvement over Intel's
offering...
>
>>AMD is nolonger a "low cost alternative". Just because AMD is a smaller
>>company(they admit that they are) doen't mean that their products are worse,
>>Borland products are much better than Microsoft ones but Microsoft is a much
>>bigger company.
>
>The Athlon is likely a superb processor.
Yes, yes it is.
> Hopefully the chipsets that are
>available for the Athlon don't leave a lot to be desired.
No, no they don't. The only thing that is left wanting is the fact
that finding an athlon M/b can be tricky, but in a few weeks it should
be much easier.
> AMD's K6
>processors have been somewhat hobbled by the low-quality of the Super
>Socket 7 chipsets and motherboards.
Hmmmm.....and the fact that they sucked compared to Intel's Mhz
equivalent cpus.....AMD has only recently developed a great FPU, and
the FPU in the K6 series' sucked.
Dionysus
offered the best price/quality ratio....my current machine has a
celeron 300a@466Mhz in it, and if in time the Athlon provides the best
$ to performance, i'll go back to it. I couldn't care less who make my
cpu, as long as it is a) cheap and b) fast....to that end i like
competition and a monopoly by Intel or AMD would be a stupid thing for
consumers to hope for.......one point against AMD is that you might
have problems wallking into your local computer store in the next few
weeks and just picking up a Athlon board and CPU....but yes AMD has
the fastest x86 processor in the world, granted.
On Sun, 29 Aug 1999 12:28:35 +0800, "Martin Kou || cybermartin"
Dionysus
being a huge design improvement over the PII/III, let alone the K6-X
series.......
On 3 Sep 1999 06:13:56 GMT, "* * Chasman" <dna...@aol.NOSPAM.com>
wrote:
<snip>
Gary Forbis
As I understand it the model is to make intentional changes which make the
competing product "non-standard" and then exploit this difference in one's
advertising. It's worked in the past for IBM, Intel, Microsoft, and many
other
industry "leaders." By putting one's competition in their place one keeps
them in reactive mode.
The rhetoric is usually something like, "Well, if you use undocumented or
non-standard features introduced by others we cannot guarantee compatibility
in our future releases."
Dionysus <mp3...@hotmail.com> wrote in message
news:37cf86dc...@news.unsw.edu.au...
Gary Dale
computer for my brother with the same configuration you have (except he is
running Windows 98 with the Xpert 128 card). To answer your questions:
- the ASUS board works well as does the AMD K6/3 450 (use one myself too)
- Check the default settings for the Ali busmaster drivers. They may not turn
on both IDE controllers (in case you have the CD-ROM connected to the second
controller)
- you have Windows 95 OSR 2. Version 2.1 is the same except it has USB drivers
with it. Try Microsoft for the drivers.
- If you just moved your disk drive, etc. from you old system to your new
system then Windows 95 may not have detected all of your hardware properly.
Your best bet is to start from scratch and re-install it and all of your other
software. Before trying that, you should run a ScanDisk (or other disk
checker) to make sure your drive hasn't lost some sectors. This is standard
advice no matter what processor or mainboard you are using.
- The AMD chip runs hot. Get a good CPU fan for it. You may also want to make
sure your case has lots of room and good ventilation. Turn on the ASUS
monitoring software (if you can get Windows to run) to monitor the temperature
for a while. Also, check the CPU voltages. Cheap power supplies can be off in
the voltages they deliver. A little high is OK but it also heats the chip up
more than necessary.
- Stay away from Intel. Their new chips have privacy issues with them.
Kelly Lu wrote:
> This is why I don't get AMD chips. Too many problems. Because of this, I
> don't actually "own" an AMD, but I would suggest trying that patch on the
> Microsoft website. I've heard countless problems (too many problems to say
> the truth) with AMD and Windows 95.
> Hope you get this solved! And next time, stick with Intel.
Andy
To run OSR2 on an AMD faster than 350MHz, you need a patch from MS.
There is some problem with timing that causes OSR2 to crash before
loading the GUI. If you later install the USB patch, you also have to
reinstall the AMD patch - the file versions get messed up.
On Fri, 22 Oct 1999 20:57:53 GMT, Gary Dale <28222...@home.com>
wrote:
Andrew Hawkesworth (ICQ #22065701)
--
M Wulfman
problem under NT so if you aren't a gamer, don't own USB devices and you
want a _way_ more stable, easier to secure OS, go NT 4.0 workstation.
Matthew
--
------------------------------------------------------------------------
(-: I hate to sound unfriendly, but to avoid confusion, the fact that my
ISP has automagically put an IP address in my header, does _NOT_
constitute an invitation to connect to my machine. Have a nice day. :-)
========================================================================
Andrew.Gill
>> This is why I don't get AMD chips. Too many problems. Because of this, I
>> don't actually "own" an AMD, but I would suggest trying that patch on the
>> Microsoft website. I've heard countless problems (too many problems to say
>> the truth) with AMD and Windows 95.
Hmmm...Have you ever used an AMD chip, because just saying they have problems with
out experiencing them is excepting everything at face value. Have you've ever read
a tabloid news paper? If you did, did you accept everything at face value, or did
you question it? Have you ever been approach by any religious group? Are you now
a member of that religion? How many religions have you been a memeber of... I
hope like any sensible inquisitive human being you do not follow the masses out of
mis-guided trust, as your post suggests.
I have had an AMD processor since the dx4-100 was released. I have never had any
problems, I currently own an AMD-225(oc 180) and a k6-2-380, and they both run
fine, infact my k6-2-380 runs perfectly stable and I use it for my cd-writing.
I've just bought (well 4month ago) a celeron 500 and that's great too, if you
never had the experience of a dx133 (oc100) kicking a p60 into the ground then
perhaps you cannot appreciate the fact that AMD is actually a great company...in
fact until the celeron I haven't owned an intel, I've had plently of experience
with any chip you might care to mention (other than the Athlon) and that includes
dec alphas and cyrix, and the chip that I have seen achieve full system crashes
most frequently was p200pro running NT, this computer crashed at least once a
week, which to me is far too much to count as stable...second comes my k6-2-380
when I oc it to 400, it crashed once every 2 weeks, so I put it at its normal
level and it hasn't crashed yet...
>> Hope you get this solved! And next time, stick with Intel.
>> Kelly Lu
The only one worth buying is the celeron, all the rest are just too expensive for
what you get.
Gary Dale wrote:
> That's crap. I've never had any problems with AMD chips. I recently set up a
> computer for my brother with the same configuration you have (except he is
> running Windows 98 with the Xpert 128 card). To answer your questions:
I'd like to second this comment, I've never had any problems with AMD chip or
non-pentium pro chips. I assume the athlon is just as good as the rest but I have
no experience of it yet...Choose your chip manufacturer based on speed and price
not on the little sticker that you can put on your case.
My advise to anyone who is listen, buy an Abit bp6 motherboard and buy a celeron
500, when win2000 comes out you've got a cheap and simple upgrade to a dual
celeron and you have a very powerful computer.
--
Andrew
LocalPinoy
It may be second fiddle in terms of sales
But rela PC gamers and integrators love AMD...
My love affair with AMD started in an accident.
I specified an Intel 386DX-25
But the tech supplied me with a AM80386DXL-40! at an even much lower
price...
What surprised me was Intel then stopped at 33
then came i486-16/25/33MHzs
And a failed intel FPU they sold as SX. those that passed... DX
Then there came the remarked issue. In an effort to duplicate the speed
attained by AMD cpus, Intel chips were set to overclocking, DX2s. If it runs
it got
remarked.
In Intel's frustration, it sued AMD for licensing rights.... to regain its
fast losing hold in sales.
Nexgen was first to venture to MMX extension by providing multimedia
processor, in effect it was the only 486 with a pentium structure.
It goes like this: if you bear with me.
add an FPU to a 386 and a data cache of 8kb It is now 486!
add to that 486 with an additional integer processing unit, with more
cache... Intel calls it pentium, it has multiprocessing/superscallar
capabilities, 2- integer units, 1 FPU, and more cache.
Nexgen had five processing units during this time, but had limited shipments
in its Nx100s, if there was.( Intel only had three.)
then the Pentium Pro?
It was a 32 bit gamble that flunked.
Now they must have revised the microcode to change the logic function of one
spare processing unit and fit it as an MX module, mount the chip in a board,
add more L2 cache, place it in a box. and renamed it Pentium II-266, Or the
Celery, the low
speed ones?
Mind you I have seen CPUs masquerading as Pentium or Celeron abound as tray
type but without the markings of a geniune Intel make.
It even runs faster than the Chip it is made to represent...
So Intel got a 370 socket for the genuinely marked 370 Chip, my first intel
a celeron366. I
mounted it on a mini board called Bridge (Slotkit). And fitted it on an
IntelEXboard (66MHz). What about an asus bridge with more features. My
bridge has some
options left out, as can be seen by vacant chip solder.
So It may cast doubt on your thinking that it is only Intel who knows their
chips?
An Athlon can be re mounted on a reconfigured bridge and mounted on an Intel
slot... Techs can do it. And subvert the conventions with data gleaned from
technical journals and kill/short some terminals on CPU ID and clock. We
can have AthlonPowered Super Pentium.
I still keep my 486DX2/DX4s, Cyrix, AMD5x86-75s, I intend to obtain
technical data for these chips, my experience as in electronics hobbyist can
be augmented with some background in programming, So I can simulate my
cellphone with my CPUs. and monitor the airwaves with some amplifiers. Who
knows. Or get a super scallar Slot one module...I know it is the
Microcode... But electrically it can be usurped.
Amd can not copy its electrical connections, the slot is patented by Intel.
It might be another MMX agreement where AMD was made to divulge its Nexgen
MX gates. So Intel could confirm if they're in the right direction
..............................
Now here is my side....to share pleas listen...(read on)
My next CPU was the DX4- AMD's achievement (actually 3 x 33.3 = 99.99 yet,
intel's then was DX2) My DX2 on acer SCSI ran conveniently faster than the
first bugged pentiums even the 75s and 90s
With the 120MHz and further did intel finally came out with a better
controller chipset.
I obtained AMD 5x86- 75/AMD-X5-133ADW (A9630CPD), and an Cyrix 6x86P150+
running almost par with my office's Pentium 166sMX
Now I have that faith in AMD and even Cyrix I am running on MII-400 and am
awaiting for the availability of AMD 750chipset boards locally.
And further Via tecnologies would not create their own chipset for Athlon if
not for its study build.
Really it is the Intel chips who has got more errata listing than AMD's
And another...
The chipset architecture is made to slow the streaming of data so to reduce
overheating...
Your intel CPU may be detected to be running at so and so speed.. but really
it is running at a much slower pace, In the controllers' speed of 33 MHz...
The board may be said to runn at 66 MHz but the PCI cards are still 33MHz.
The CPU runs internally at 4x but that translates to 8 waits per cycle.
All it does is predict. And predict it does until the other proc-unit spews
out the same data.
It is still 80816 logic made to run faster.
AMD has a RISC unit translating to CISC code at the gate. I runs therefore
more efficiently inside or else the translation with superscallar will be
garbage. And no divide error permissible as with pentiums, where there would
be deviations in the internal clock and mismatch of cycles caused by
overheating/overdrive. Pentium logic therefore should be more efficient in
its native 86 mode, but it continues to have stepping and errata text issued
with every batch of shipment.
(But we know all logic chips have this, lol)..
I love AMD, it challenged Intel to excel. In that I benefitted myself with
choices. Suitable for my need. Thanks to you, Folks.
I have integrated a lot of Pentium IIs and a few Pentium IIIs, but the speed
seem the same..It is always 29 minutes for a win98 intall/setup.
But with AMD and Cyrix, I have varying degrees corresponding to it's latent
speed.
And all the CPus I have had been over clocked.
This CPU I am using, a Celery 366 runs at 410 at 68.7board
clock(IntelEX-983DTomato). My other Cyrix MII-400 is running at 433, with
DIP settings (Octek Via 598MVP Board).
My AMDX4100 and 5x86 are with my kids in their Jumpstart Sessions.
One Pentium III in my sister's house has the outfitter yet to be
installed...
She is not willing to do it, (the 666?) A serial number for her name....
Funny, but could be true... a pre cursor.... Intel are you listening?
It has to happen. These are but milestones... Then the end will come.
Andrew.Gill <Andre...@durham.ac.uk> wrote in message
news:3817C19C...@durham.ac.uk...