Jason Dean Malone (jdmalone) wrote:
: System config:
: ASUS PVI486-SP3 mainboard
: 256k interleaved cache
: Award 4.50PG BIOS, version 304
: Hercules Dynamite Power VESA, 2 MB
: 8 MB RAM
:
ASUS VLISV2GX4 256KB cache STB LightSpeed VL 2MB (same chipset as herc) 16MB
:
: Sysinfo 8.0 PC Config 8.03 3d-Bench
: Cx5x86-120 ~260-380 (?) 19250% 90.9
: Pentium-90 288.3 19850% 83.3
: Pentium-100 316.9 24350% 100.0
: AMD 5x86-133 288.7 18050% 90.9
AMD 5x86-150(3x50) 324.7 100.0
: AMD 5x86-160 346.1 23550% 100.0
: Pentium-120 380.2 28850% 111.1
my next attempt will be to try a Cyrix 5x86 at 150MHz....
Okay,
System config:
ASUS PVI486-SP3 mainboard
256k interleaved cache
Award 4.50PG BIOS, version 304
Hercules Dynamite Power VESA, 2 MB
8 MB RAM
Sysinfo 8.0 PC Config 8.03 3d-Bench
Cx5x86-120 ~260-380 (?) 19250% 90.9
Pentium-90 288.3 19850% 83.3
Pentium-100 316.9 24350% 100.0
AMD 5x86-133 288.7 18050% 90.9
AMD 5x86-160 346.1 23550% 100.0
Pentium-120 380.2 28850% 111.1
Pentium scores (90, 100 MHz) taken in:
Intel Premeire II MB, Neptune chipset, 256k cache. Diamond Stealth 64
Pentium 120 score:
Taken in my other PC w/ Triton chipset, 512k PB cache, Hercules
Stingray/64
Cyrix 5x86 scores:
Averages taken from various sources
So, If you're looking for a good "upgrade" processor for
your 486, look no further (assuming your MB could handle
the O.C.) The AMD-160 definitely gives both the Pentium-90
and Pentium-100 a run for their money. (Best 50$ i've EVER
spent)
>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>and i'm pleased to say that it's FAASSSSSTTTTT!
>It definitely outperforms most P-90's, and comes
>close to the Pentium-100 CPU.
>Thought i'd post the CORRECT benchmarks of the
>AMD-160 chip to clear up some confusion running
>around. (yes, it's faster than the 5x86-120---hands down.)
>
>Okay,
>
>System config:
>ASUS PVI486-SP3 mainboard
>256k interleaved cache
>Award 4.50PG BIOS, version 304
>Hercules Dynamite Power VESA, 2 MB
>8 MB RAM
>
>
> Sysinfo 8.0 PC Config 8.03 3d-Bench
>Cx5x86-120 ~260-380 (?) 19250% 90.9
>
>Pentium-90 288.3 19850% 83.3
>
Holy Shit is right!!! And just when I was thinking of getting a new
motherboard. Could you tell me exactly what you did to overclock the
chip to this speed? Did you boost the speed of the ISA bus above 33?
What speed is the base speed of the processor? I'm knowledgable about
the BIOS and jumper settings, but I'm kind of new to the overclocking
world. What can I do with my present DX2/66? Any info you can send me
will really be appreciated, and I'll let you know my results!!!
Regards,
Greg Hughes
Both the Cyrex and AMD 5x86 chips are FAR superior to ANYTHING ! that
pathetic Intel DREAMS of putting out !
However there are some things you need to know, I will share with you my
experiences with the 5x86 series of processors that I have had in the
past 2 months:
First off :
If you are purchasing chips that come with the motherboards, as a
package, DO NOT trust the jumper setting EVEN IF THEY ARE SILKSCREENED ON
THE BOARD. I have had quite a nightmare with this , but with a little
patience , I was able to safly pump up both AMD and Cyrex to well iver
140 mhz. But remember , you need to have a jumper selectable MB, dont try
it with MB that use one oscillator, it will not work.
If you want to overpump a 486 66 then as a maximum, use a 80 mhz
oscillator , and by all means use a heat sing with a respectable set of
balls.
Of particular interest to me is overclocking AMD 386 DX 40 chips. I have
pumped up these babies to well over 66 MHZ and I burn them in by running
them for about 1 week , non stop WITHOUT a heat sink. I have yet to see
one fail.
Fabo
fab...@li.net
>Of particular interest to me is overclocking AMD 386 DX 40 chips. I
have
>pumped up these babies to well over 66 MHZ and I burn them in by
running
>them for about 1 week , non stop WITHOUT a heat sink. I have yet to
see
>one fail.
>Fabo
>fab...@li.net
Hi... just curious: It sounds like you do this A LOT, and not just
for yourself. Do you sell these overclocked motherboard/CPU's?
The only concern I'd have is that your week long burn-in (which seems
like a lot of effort to save a few bucks, btw) might not detect a
problem with a seldom used opcode combination.
I am curious to know what is it about your business that leads to you
doing so much over-clocking? Do people actually _buy_ pre-overclocked
systems?? Over-clocking is a fun way to feel like your getting more
bang for your buck, but I'd think that's about the extent of it. I'd
certainly never _buy_ an overclocked system.
Chris Bennett (finger 'benn...@MCS.COM' for WWW, FTP, PGP, etc.)
"If, in reading this, you are uncertain as to whether a specific statement
is meant seriously or not, simply apply this rule of thumb: If the statement
makes you consider filing a lawsuit, I was kidding. Ha ha!" - Dave Barry
>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>and i'm pleased to say that it's FAASSSSSTTTTT!
>It definitely outperforms most P-90's, and comes
>close to the Pentium-100 CPU.
>Thought i'd post the CORRECT benchmarks of the
>AMD-160 chip to clear up some confusion running
>around. (yes, it's faster than the 5x86-120---hands down.)
>Okay,
>System config:
>ASUS PVI486-SP3 mainboard
>256k interleaved cache
>Award 4.50PG BIOS, version 304
>Hercules Dynamite Power VESA, 2 MB
>8 MB RAM
> Sysinfo 8.0 PC Config 8.03 3d-Bench
>Cx5x86-120 ~260-380 (?) 19250% 90.9
>Pentium-90 288.3 19850% 83.3
>Pentium-100 316.9 24350% 100.0
Can you give me the name of your vendor?
I'm interested on buying AMD-160 for $50
[Posted with Agent .99d. For info, email agent...@forteinc.com.]
>On 16 Feb 1996 03:17:35 GMT, Jason Dean Malone <jdmalone> wrote:
>
>>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>>and i'm pleased to say that it's FAASSSSSTTTTT!
>>It definitely outperforms most P-90's, and comes
>>close to the Pentium-100 CPU.
>>Thought i'd post the CORRECT benchmarks of the
>>AMD-160 chip to clear up some confusion running
>>around. (yes, it's faster than the 5x86-120---hands down.)
>>
>>Okay,
>>
>>System config:
>>ASUS PVI486-SP3 mainboard
>>256k interleaved cache
>>Award 4.50PG BIOS, version 304
>>Hercules Dynamite Power VESA, 2 MB
>>8 MB RAM
Snip...
Hello Jason - Son of all overclockers....
Ooooooohhh YEEES ! IIIII love this !!
Please tell me how you did this (step by step)
I have the exact same motherboard as You !
I´m getting stomach-cramps of pure breathless excitement !!!
ps. Is this safe ? (my money are all gone!)
Regards Kenneth R
(oh my stomach...)
<CUT>
> Holy Shit is right!!! And just when I was thinking of getting a new
> motherboard. Could you tell me exactly what you did to overclock the
> chip to this speed? Did you boost the speed of the ISA bus above 33?
The ISA-Bus only runs at 8 MHz (okay, tuned it will run al 11 MHz too..)
Question bus-speed..... :
Thats a think, I would also be interested in..
Overclocking a 133 to 160 means running with 40 MHz external speed...
So, the PCI-Bus is also running at 40 MHz or does the Mainboard spilts
to speed to 20 MHz (??).
Running the PCI-Bus (with my Winner2000AVI-Graphics-Card) at 40 MHz
might cause a lot of trouble ??
Bus-speed tunig is always a little bit tricky, because of the exact
timings , the PCI-bus needs...
And nobody can really measure the PCI-Bus Clocks.....
And comments are welcome ...
Greets
Michael
> Regards,
> Greg Hughes
>
> ghug...@ix.netcom.com
>Hi
> Take it from a guy that builds custom computers. and also sells
>at computer shows.
Uh oh...another "expert." Based upon the below post, you should think more
than twice before listening to, or buying, anything from this clown.
>Both the Cyrex and AMD 5x86 chips are FAR superior to ANYTHING ! that
>pathetic Intel DREAMS of putting out !
Superior to the i486 chips, sure. However, not superior to ANYTHING intel
produces (if money were not a factor, would you rather have a 5x86 or a
p5-166?)
>However there are some things you need to know, I will share with you my
>experiences with the 5x86 series of processors that I have had in the
>past 2 months:
>First off :
>If you are purchasing chips that come with the motherboards, as a
>package, DO NOT trust the jumper setting EVEN IF THEY ARE SILKSCREENED ON
>THE BOARD. I have had quite a nightmare with this , but with a little
>patience , I was able to safly pump up both AMD and Cyrex to well iver
>140 mhz. But remember , you need to have a jumper selectable MB, dont try
>it with MB that use one oscillator, it will not work.
I'll agree with faulty screens on a few boards. However, I've also seen faulty
manuals. What should a person trust, pray tell? Just randomly move jumpers?
And if you're going to sell "Cyrex," at least spell it correctly.
140 mhz??? Most 486 boards are selectable for 25, 33, 40 and/or 50 mhz. 140
mhz is a mystery multiple of one of these clock rates.
>If you want to overpump a 486 66 then as a maximum, use a 80 mhz
>oscillator, and by all means use a heat sing with a respectable set of
>balls.
I love those heat sings...they are music to my ears!
>Of particular interest to me is overclocking AMD 386 DX 40 chips. I have
>pumped up these babies to well over 66 MHZ and I burn them in by running
>them for about 1 week , non stop WITHOUT a heat sink. I have yet to see
>one fail.
You are either insane, a liar, or really stupid...I vote the latter. Burn them
in...right. First, practically all 386 boards are those single clock boards
you mentioned earlier. To change the clock, you must swap out the mod-osc can
which runs at 2x the cpu clock. This certainly can be done, with some success
(especially 50 to 66 mhz and 33 to 80). I've tried to run a 100mhz in mine
(for a 386dx50) with no success. While it is possible the cpu will run that
fast, I'm pretty sure the board itself won't...it certainly won't run with a
130+mhz osc. Besides, where are you finding such parts? I've never seen a ttl
mod osc that runs at over 100mhz (I'd love to find a 88 or 90 mhz can, but the
only thing I've found faster than an 80 is the 100).
There is absolutely no 386/486 board which runs a cpu at an external clock
greater than 50 mhz. Period.
You're probably running your 386dx40 at 33 rather than 66. No wonder it's
stable. That burn-in technique of yours is also a joke.
Once again, a computer technician/vendor/expert who couldn't find his ass with
both hands. He knows even less about computers than that esquire@io knows
about law!!!
>Fabo
>fab...@li.net
>ghug...@ix.netcom.com (Greg Hughes) wrote:
>>On 16 Feb 1996 03:17:35 GMT, Jason Dean Malone <jdmalone> wrote:
>>
>>>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>>>and i'm pleased to say that it's FAASSSSSTTTTT!
>>>It definitely outperforms most P-90's, and comes
>>>close to the Pentium-100 CPU.
>>>Thought i'd post the CORRECT benchmarks of the
>>>AMD-160 chip to clear up some confusion running
>>>around. (yes, it's faster than the 5x86-120---hands down.)
>>>
>>>Okay,
>>>
>>>System config:
>>>ASUS PVI486-SP3 mainboard
>>>256k interleaved cache
>>>Award 4.50PG BIOS, version 304
>>>Hercules Dynamite Power VESA, 2 MB
>>>8 MB RAM
>>>
>>>
>>> Sysinfo 8.0 PC Config 8.03 3d-Bench
>>>Cx5x86-120 ~260-380 (?) 19250% 90.9
>>>
>>>Pentium-90 288.3 19850% 83.3
>>>
>>Holy Shit is right!!! And just when I was thinking of getting a new
>>motherboard. Could you tell me exactly what you did to overclock the
>>chip to this speed? Did you boost the speed of the ISA bus above 33?
: Question bus-speed..... :
: Thats a think, I would also be interested in..
: Overclocking a 133 to 160 means running with 40 MHz external speed...
: So, the PCI-Bus is also running at 40 MHz or does the Mainboard spilts
: to speed to 20 MHz (??).
: Running the PCI-Bus (with my Winner2000AVI-Graphics-Card) at 40 MHz
: might cause a lot of trouble ??
: Bus-speed tunig is always a little bit tricky, because of the exact
: timings , the PCI-bus needs...
: And nobody can really measure the PCI-Bus Clocks.....
I've got my AMD 486 running at 160 MHz (very stable) with the PCI bus
running at 40 MHz. On my board the PCI bus speed is selectable through
the BIOS settings. BTW, I'm using a #9 Vision 330 video card.
--
******************************************************************************
Raymond Fellers | | | |\ | | | \ /
The Saykally Research Group | |__ | | \| |__| / \
Laser Spectroscopy/VdW Clusters |
r...@uclink3.berkeley.edu | http://www.zoom.com/~rsf
******************************************************************************
: Sysinfo 8.0 PC Config 8.03 3d-Bench
: Cx5x86-120 ~260-380 (?) 19250% 90.9
:
: Pentium-90 288.3 19850% 83.3
:
: Pentium-100 316.9 24350% 100.0
:
: AMD 5x86-133 288.7 18050% 90.9
:
: AMD 5x86-160 346.1 23550% 100.0
:
: Pentium-120 380.2 28850% 111.1
:
[more stuff deleted]
For one thing -- use a REAL benchmark. SI 8.0 runs COMPLETELY out
of the internal cache (turn of the external cache -- it won't make any
difference). From what I see, so does the PC-Config. 3d-bench is a total
joke -- I've got the same computer and the same SI benchmark but I get 67
FPS. Why? Because I have a slow video card. Why don't you post something
useful, like a PCBENCH Cpumark or something. Just to get things started I
get a 150 with my AMD5x86 at 50x3 (this is pretty slow, BTW, but my old
486-33 got 42 so...)
Overall I do think the processor is good - very good memory
performance and good overall CPU performance; just don't compare it
without using something that will give real-world results. The other
thing is that not all the CPUs will do 160Mhz (mine won't, even with all
the wait states on -- runs fine in DOS, but crashes in Windows and
Autocad). There's a lot more to the computer than just the processor.
:
: So, If you're looking for a good "upgrade" processor for
: your 486, look no further (assuming your MB could handle
: the O.C.) The AMD-160 definitely gives both the Pentium-90
: and Pentium-100 a run for their money. (Best 50$ i've EVER
: spent)
:
Oh, one other thing -- the FPU is REALLY slow compared to the
Pentium ones. About 50%-60% as fast as a P-90. Good if you never use it
--otherwise don't waste your money.
Just my $.02
L8r-
-Lance
ASUS PVI-486SP3?
1) Open up ur casing.
2) Look for a small chip between the VL slot and the cache ram.
3) If the chip reads VIA VT8228 then set jumper
JP25 JP26 JP27 JP28
from 33MHz 1&2 2&3 2&3 1&2
to 40MHz 1&2 1&2 1&2 2&3
else if chip reads AV9155
JP25 JP26 JP27 JP28
from 33MHz 1&2 1&2 2&3 1&2
to 40MHz 1&2 2&3 1&2 1&2
That's it.
Any one knows the setting for 50MHz, it is not in my manual.
--
~Frisky Kitty~
!*% T-3..............T-2.............T-1.............BOOM.....................
As most of you that have been following this thread know, and those of
you that have experimented with said chips KNOW. What I have spoken here
and the advice I have given to you is for all practical purposes , TRUE.
Ok, that out of the way. It apears that this gentlemen has a VAST
knowledge of , well, nothing, well, I shouldnt say that, he seems to me
to have a keen eye for spelling errors.
It is JUST THESE people that the eager to learn consumer needs to watch
out for.
Ill state some cases of the uniformed claimes that he vomited SOOO
irresponsibly.
He apparantly KNOWS that Intel is the fastest chip around
( This proves my theory that the dong ding dong ding sound at the end of
the intel comercials is in actuality a subliminal brainwashing technique
that works only on Superintelligent specimens like this fella)
He displays the TYPICAL " I KNOW you're wrong , but Ill challenge you to
explain it in words my larvae brain can handle" ( That way, I can, if it
sinks in to my limited intelligence, go and shoot my mouth off to other
people , like I really know what Im talking about)
Look, I can go on, but those of you that understood what I was saying,
know these things to be true.
Those that dont.... Well asking is always good, but make sure that you
get a second opinion. That goes for any advice that I give you also.
As for this guy, well, hey if he is ever in NY and would like a lesson or
two on system configuration, Hey Ill do whatever I can to illuminate a
dull mind.
I ask all who have been following and contributing to this tred, that you
reply to this, dont flame the poor guy, TEACH !
Later !
Fabo
fab...@li.net
/////////////// stuff deleted ////////////////
How about taking this rambling babble off line.
>Hi:
>Below is an example of " There is no one as well informed as the mis
>informed"
>As most of you that have been following this thread know, and those of
>you that have experimented with said chips KNOW. What I have spoken here
>and the advice I have given to you is for all practical purposes , TRUE.
I have never seen any reports, except for yours, of overclocking a 386dx40,
particularly to 66 mhz. The fact is that the motherboards are simply incapable
of running at such speeds. You are likely underclocking your board, but are
simply too stupid to realize it.
>Ok, that out of the way. It apears that this gentlemen has a VAST
>knowledge of , well, nothing, well, I shouldnt say that, he seems to me
>to have a keen eye for spelling errors.
>It is JUST THESE people that the eager to learn consumer needs to watch
>out for.
Well, I'm not trying to sell anyone anything. You're the one spreading bad
advice and (perhaps) hosing nice consumers downin your role as computer
expert and vendor. As for spelling errors, I don't flame common typos...but
for a "vendor" to grossly misspell his wares (the "i" isn't too close to the
"e" on your keyboard, is it?) is an open invitation.
>Ill state some cases of the uniformed claimes that he vomited SOOO
>irresponsibly.
Uniformed? My vomited "claimes" are naked as a jaybird, fabio.;)
>He apparantly KNOWS that Intel is the fastest chip around
>( This proves my theory that the dong ding dong ding sound at the end of
>the intel comercials is in actuality a subliminal brainwashing technique
>that works only on Superintelligent specimens like this fella)
My, you've developed theories about the effects of the mass media. Quite the
budding George Gerbner, aren't we? Of course, you didn't read my post very
well. I like the AMD and Cyrix 486 clones (well,except that those Cyrix 486
dlc/slc things were a piece of shit!). PC Mag now reports a faster
winstone for a Cyrix 6x86 p166+ than any tested p5-166. However, the fastest
machines generally on the market are all Pentiums, not Cyrix, AMD, nor Nexgen.
I'm no intel lover, but intel products are not automatically crap.
>He displays the TYPICAL " I KNOW you're wrong , but Ill challenge you to
>explain it in words my larvae brain can handle" ( That way, I can, if it
>sinks in to my limited intelligence, go and shoot my mouth off to other
>people , like I really know what Im talking about)
> Look, I can go on, but those of you that understood what I was saying,
>know these things to be true.
No one knows how you oc a 386 to 66 mhz. Please enlighten us. Tell us how many
wait states you need to put in for the cache and main memory? Do we change
jumpers or the mod-osc can? C'mon fabioboy, we larva brains want to know!
>Those that dont.... Well asking is always good, but make sure that you
>get a second opinion. That goes for any advice that I give you also.
Fabio don't know diddly. If you ask for a second opinion about his advice
(note that he is unable to tell anyone how to oc the 386), they would likely
laugh at you and tell you fab is one dumb bunny.
>As for this guy, well, hey if he is ever in NY and would like a lesson or
>two on system configuration, Hey Ill do whatever I can to illuminate a
>dull mind.
Please enlighten me, oh great shaven-chested blonde-tressed calendar boy
fabio. Tell all of us with 386 machines how to "overpump." I don't know
everything, but I do know that I can lay claim (at least until someone posts a
better score) to having THE FASTEST 386 IN THE ENTIRE WORLD. That's right,
consult the pcbench 9.0 benchmark list, look towards the end of the list (hey,
it's still just a 386!), and the highest ranking 386 is mine. I've actually
tweaked it a tad higher since then, getting a dosmark of about 215 (I'll have
to double check the exact number when I get home). I've tried oc'ing the
machine to a 50 mhz, but it would hang, sometimes on boot and sometimes
shortly after starting. If someone can steer me to a source for a 90mhz mod
osc, I'll be a happy man.
>I ask all who have been following and contributing to this tred, that you
>reply to this, dont flame the poor guy, TEACH !
Flame away...just do a better job than this mental midget. And teach as well.
Let's all have faster machines. If fabioboy can tell anyone how to make a 386
run at 66 mhz, I'll publicly apologize and hail his technical ingenuity. Until
then, I think he's a halfwit.
>Later !
>Fabo
>fab...@li.net
*SNIP*
Hey fabio baby, no need to cc me a copy of your little inanities. If they're
posted, I'll see them.
Lionel Hutz, Esq. THE MAN WITH THE FASTEST 386 IN THE WORLD
HA!
<snip>
>Overclocking a 133 to 160 means running with 40 MHz external speed...
>So, the PCI-Bus is also running at 40 MHz or does the Mainboard spilts
>to speed to 20 MHz (??).
>Running the PCI-Bus (with my Winner2000AVI-Graphics-Card) at 40 MHz
>might cause a lot of trouble ??
<snip>
My understanding of the PCI Architecture is that it runs at a constant
33MHz, regardless of what speed the system is running. I believe this
applies whether the base speed of the system is 25, 33, 40, 50, 60 or
66 (less any clock multiplying done by the CPU)
Thanks
Greg Pinto
(snip)
>Ok, that out of the way. It apears that this gentlemen has a VAST
>knowledge of , well, nothing, well, I shouldnt say that, he seems to me
>to have a keen eye for spelling errors.
(snip)
>He apparantly KNOWS that Intel is the fastest chip around
You mean he can't spell words like apparently? :")
Joe
>Hi
> Take it from a guy that builds custom computers. and also sells
>at computer shows.
>Both the Cyrex and AMD 5x86 chips are FAR superior to ANYTHING ! that
~~~~~
!?!?! You know what you are talking about, dont you?
>pathetic Intel DREAMS of putting out !
>However there are some things you need to know, I will share with you my
>experiences with the 5x86 series of processors that I have had in the
>past 2 months:
>First off :
>If you are purchasing chips that come with the motherboards, as a
>package, DO NOT trust the jumper setting EVEN IF THEY ARE SILKSCREENED ON
>THE BOARD. I have had quite a nightmare with this , but with a little
>patience , I was able to safly pump up both AMD and Cyrex to well iver
>140 mhz. But remember , you need to have a jumper selectable MB, dont try
~~~~~~~
!?!?!?! Guy, you are very funny.
>it with MB that use one oscillator, it will not work.
>If you want to overpump a 486 66 then as a maximum, use a 80 mhz
>oscillator , and by all means use a heat sing with a respectable set of
~~~~~~~~~~~
!?!?!?!?!?! This will really PUMP UP the chip. Better get a 40mhz
oscillator. (especially for you: 2x40=80)
>balls.
>Of particular interest to me is overclocking AMD 386 DX 40 chips. I have
>pumped up these babies to well over 66 MHZ and I burn them in by running
~~~~~~~~~~~~~~
You are a LAMER.
>them for about 1 week , non stop WITHOUT a heat sink. I have yet to see
>one fail.
>Fabo
>fab...@li.net
>ghug...@ix.netcom.com (Greg Hughes) wrote:
>>On 16 Feb 1996 03:17:35 GMT, Jason Dean Malone <jdmalone> wrote:
>>
>>>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>>>and i'm pleased to say that it's FAASSSSSTTTTT!
>>>It definitely outperforms most P-90's, and comes
>>>close to the Pentium-100 CPU.
>>>Thought i'd post the CORRECT benchmarks of the
>>>AMD-160 chip to clear up some confusion running
>>>around. (yes, it's faster than the 5x86-120---hands down.)
>>>
>>>Okay,
>>>
>>>System config:
>>>ASUS PVI486-SP3 mainboard
>>>256k interleaved cache
>>>Award 4.50PG BIOS, version 304
>>>Hercules Dynamite Power VESA, 2 MB
>>>8 MB RAM
>>>
>>>
>>> Sysinfo 8.0 PC Config 8.03 3d-Bench
>>>Cx5x86-120 ~260-380 (?) 19250% 90.9
>>>
>>>Pentium-90 288.3 19850% 83.3
>>>
>>Holy Shit is right!!! And just when I was thinking of getting a new
>>motherboard. Could you tell me exactly what you did to overclock the
>>chip to this speed? Did you boost the speed of the ISA bus above 33?
>>What speed is the base speed of the processor? I'm knowledgable about
>>the BIOS and jumper settings, but I'm kind of new to the overclocking
>>world. What can I do with my present DX2/66? Any info you can send me
>>will really be appreciated, and I'll let you know my results!!!
>>
>>Regards,
>>Greg Hughes
>>
>>ghug...@ix.netcom.com
Man liest sich, /\/\artin.
--------------------------------------------------
email: martin....@post.rwth-aachen.de
--------------------------------------------------
>fab...@li.net schrieb:
Und speziell fuer Dich, Martin, auf aelteren MBs wo man noch keine Jumper
zum Umstellen der externen Taktfrequenz (Stichwort:
Clockspeedgenerator) benutzt hat. Wurde IMMER ein Oszillator mit der
DOPPELTEN externen Frequenz der CPU benutzt. Also wer ist hier der
Lamer? ;-))
Sorry to the others. But I had to write in German, because I wanted Martin
to understand what I have written. It is not that interesting for the rest.
>>balls.
>>Of particular interest to me is overclocking AMD 386 DX 40 chips. I have
>>pumped up these babies to well over 66 MHZ and I burn them in by running
> ~~~~~~~~~~~~~~
>You are a LAMER.
Inform yourself before you write stupid things and insult people!
>>them for about 1 week , non stop WITHOUT a heat sink. I have yet to see
>>one fail.
>>Fabo
>>fab...@li.net
>Man liest sich, /\/\artin.
I hope not, if it will be such a rubbish again!
>--------------------------------------------------
>email: martin....@post.rwth-aachen.de
>--------------------------------------------------
Bye,
Dirk
Oh, I like it to reply to my own postings ...
Okay. It is definitly WRONG, the the PCI-bus always runs at 33 MHz.
(sorry) If zou are using an P75, the bus is only running at 25 MHz.
(i'm sure !). So the PCI-bus-speed depends an the systemspeed, divied by
2 or 3 or ...
Bye
MICHAEL
>Holy Shit is right!!! And just when I was thinking of getting a new
>motherboard. Could you tell me exactly what you did to overclock the
>chip to this speed? Did you boost the speed of the ISA bus above 33?
>What speed is the base speed of the processor? I'm knowledgable about
>the BIOS and jumper settings, but I'm kind of new to the overclocking
>world. What can I do with my present DX2/66? Any info you can send me
>will really be appreciated, and I'll let you know my results!!!
With your current DX2-66, you can successfully overclock it to 80Mhz.
I've done just that with an AMD DX2-66.
Yousuf Khan
Not exactly. The PCI spec allows any speed from 0 to 33MHz (a newer
extension in 2.1 allows 0-66 MHz, but nobody in the PC world has chips
for this yet). Most 486 boards seem to be running the PCI bus at the
memory speed, tho some might be dividing 40 MHz by 2. Most Pentium
motherboards run the PCI at half the memory speed (ie: 30 MHz for
60 MHz memory busses on P60/90/120/150, 33 MHz for 66 MHz memory
busses on P66/100/133/166, 25 MHz for 50 MHz memory bus on P75).
However the PCI architecture *allows* for asynchronous PCI bus clocks,
such that a P120 could runs a 33 MHz PCI bus for example. I don't
think any of the common chipsets (SIS, Triton, etc) handle this yet, though.
So you are right that the PCI architecture CAN run at a constant rate
independent of processor speeds; but wrong in that all the PC real world
systems I know of use synchronous (locked to CPU clock) implementations
so far.
Zhahai
>>world. What can I do with my present DX2/66? Any info you can send me
>>will really be appreciated, and I'll let you know my results!!!
> With your current DX2-66, you can successfully overclock it to 80Mhz.
> I've done just that with an AMD DX2-66.
You can also use it at 100MHz,but you have to set the voltage to 4V (3.3V
isn't enough for it to run at 100)
I did it with my CxDX2/66,and it worked without problems (but because of the
shit motherboard,I had to set up all waitstates to maximum,so the video
speed rather went down).I wasn't as heat,than an intel4/100 at the same
clock (after about 30 minutes)
stanley
>However the PCI architecture *allows* for asynchronous PCI bus clocks,
>such that a P120 could runs a 33 MHz PCI bus for example. I don't
>think any of the common chipsets (SIS, Triton, etc) handle this yet, though.
>So you are right that the PCI architecture CAN run at a constant rate
>independent of processor speeds; but wrong in that all the PC real world
>systems I know of use synchronous (locked to CPU clock) implementations
>so far.
> Zhahai
My ASUS P54SP4 allows me to run at either 1/2 or 2/3 of the CPU clock
speed. So my P100 PCI bus runs at either 33 or 44 MHz.
Cards that seem to work fine at 44Mhz:
Matrox Millennium
Adaptec 2940W
CMD Technologies PCI-to-IDE interface on the motherboard.
Cards that don't work:
Linksys PCI ethernet (with DEC chipset)
DOOM 2 benchmark increased about 15% with the 33% bus speed boost.
My Cyrix Cx5x86 system which has a 40 MHz CPU clock (x3) also uses an
asynchronous PCI clock.
---John "33MHz is fast enough for me" Shalamskas
John Shalamskas http://www.mlh.com/ jo...@mlh.com
dba MLH Consulting phone (808) 521-3141
1620 Keeaumoku St #701 pager 288-2799
Honolulu, HI 96822 "Lucky I live Hawaii!" FAX/Data 534-0579
Just my 4 cents worth (i overclocked my 2 cents..)
Chad
What a Jurk
>Welp, I've successfully overclocked the AMD-133 to 160 MHz,
>and i'm pleased to say that it's FAASSSSSTTTTT!
>It definitely outperforms most P-90's, and comes
>close to the Pentium-100 CPU.
>Thought i'd post the CORRECT benchmarks of the
>AMD-160 chip to clear up some confusion running
>around. (yes, it's faster than the 5x86-120---hands down.)
>Okay,
>System config:
>ASUS PVI486-SP3 mainboard
>256k interleaved cache
>Award 4.50PG BIOS, version 304
>Hercules Dynamite Power VESA, 2 MB
>8 MB RAM
> Sysinfo 8.0 PC Config 8.03 3d-Bench
>Cx5x86-120 ~260-380 (?) 19250% 90.9
>Pentium-90 288.3 19850% 83.3
>Pentium-100 316.9 24350% 100.0
>AMD 5x86-133 288.7 18050% 90.9
>AMD 5x86-160 346.1 23550% 100.0
>Pentium-120 380.2 28850% 111.1
>Pentium scores (90, 100 MHz) taken in:
>Intel Premeire II MB, Neptune chipset, 256k cache. Diamond Stealth 64
>Pentium 120 score:
>Taken in my other PC w/ Triton chipset, 512k PB cache, Hercules
>Stingray/64
>Cyrix 5x86 scores:
>Averages taken from various sources
>So, If you're looking for a good "upgrade" processor for
>your 486, look no further (assuming your MB could handle
>the O.C.) The AMD-160 definitely gives both the Pentium-90
>and Pentium-100 a run for their money. (Best 50$ i've EVER
>spent)
where did u get your 5x86 (dx4/133) with Motherboard at? iam looking to get
one of them, but want to know a good place and price to get it at!
so if u can tell me where u got it at and i will look in to it, cause iam
stuck with a 486dx/33 right now (WAY to SLOW!).. so thanks for the info..
I need the 5x86(dx4/133) and the MotherBoard u were talking about in this
post..
Whjat speed chips werew you using? Can you get by with 4x36-70, or do
they have to be 60ms?
Steven
Brett "waiting for AMD to ship a 166Mhz so I can clock it at 200" Wooldridge
bre...@io.com
Gino
Well, you got your answer when your system froze on bootup at the
120 MHz clock speed (assuming you're using a cooling fan). At least
you got 100 MHz out of it...
--
Mike
My opinions, not Argonne's...
>Kenneth (k...@imbmed.ou.dk) wrote:
>: Please tell me how you did this (step by step)
>: I have the exact same motherboard as You !
> ASUS PVI-486SP3?
> 1) Open up ur casing.
> 2) Look for a small chip between the VL slot and the cache ram.
> 3) If the chip reads VIA VT8228 then set jumper
> JP25 JP26 JP27 JP28
> from 33MHz 1&2 2&3 2&3 1&2
> to 40MHz 1&2 1&2 1&2 2&3
> else if chip reads AV9155
>
> JP25 JP26 JP27 JP28
> from 33MHz 1&2 1&2 2&3 1&2
> to 40MHz 1&2 2&3 1&2 1&2
> That's it.
> Any one knows the setting for 50MHz, it is not in my manual.
Just take the settings for 25MHz and set JP25 to 2&3. That's all.
>--
>~Frisky Kitty~
>!*% T-3..............T-2.............T-1.............BOOM.....................
Bye,
Dirk
eeeeeoooo!
Wrong!
PCI bus runs at different speeds according to the speed of your Pentium
( or 486) processor. Most cards are (probably) designed for 33 MHz and
below as this is the fastest speed Intel boards take them to.
>Excuse me, but I don't quite understand the rationale here..
>The P90 or P100 run at 90 MHz and 100 MHz internal speed respectively,
>but the Cyrix and AMD 5x86 chips need to run 120, 133, or 160 MHz
>to match a P100 and you think this is *good*?? If I liquid nitrogen
>cool a 386 or 486 and overclock it to 500 MHz, it will outperform
>a P100 too.. All this proves is that the P-100 has a superior
>architecture to the 5x86 since it can provide the same level of
>performance at lower clock speeds.. On the other hand if the results
>had been that the 5x86 at 90 or 100 MHz ran faster than Pentium
>at 133 or 160 MHz, then that would be something to talk about...
Chris --
I don't think that anyone is trying to claim that the AMD 5x86-P75 is
a "better" CPU than a true Pentium. The point is that for individuals
who currently have 486 motherboards, the AMD and Cyrix 5x86 CPUs
provide a level of performance that compares to a P75, P90 or whatever
for a far smaller investment.
If you are in the market for a new system, by all means get a Pentium.
However, for less than $120 (the CPU plus a voltage regulator), I can
breathe new life into an older 486 computer with an AMD 5x86.
Jim Stansell
Institute for Public Policy and Social Research
321 Berkey Hall
Michigan State University
East Lansing, MI 48824
Ditto.
Andrew
I'm considering overclocking my AMD 133mhz PCI setup to either 150 or 166mhz.
If I go the 150 route, the external bus will run at 40mhz vs 33mhz at
166mhz. Which configuration will provide me with the overall best
performance? Will moving from 33 to 40mhz external speed make a noticable
difference?
Thanks,
-Frank-
<PC config deleted>
: Sysinfo 8.0 PC Config 8.03 3d-Bench
: Cx5x86-120 ~260-380 (?) 19250% 90.9
: Pentium-90 288.3 19850% 83.3
: Pentium-100 316.9 24350% 100.0
: AMD 5x86-133 288.7 18050% 90.9
: AMD 5x86-160 346.1 23550% 100.0
: Pentium-120 380.2 28850% 111.1
: Pentium scores (90, 100 MHz) taken in:
: Intel Premeire II MB, Neptune chipset, 256k cache. Diamond Stealth 64
: Pentium 120 score:
: Taken in my other PC w/ Triton chipset, 512k PB cache, Hercules
: Stingray/64
: Cyrix 5x86 scores:
: Averages taken from various sources
: So, If you're looking for a good "upgrade" processor for
: your 486, look no further (assuming your MB could handle
: the O.C.) The AMD-160 definitely gives both the Pentium-90
: and Pentium-100 a run for their money. (Best 50$ i've EVER
: spent)
Excuse me, but I don't quite understand the rationale here..
The P90 or P100 run at 90 MHz and 100 MHz internal speed respectively,
but the Cyrix and AMD 5x86 chips need to run 120, 133, or 160 MHz
to match a P100 and you think this is *good*?? If I liquid nitrogen
cool a 386 or 486 and overclock it to 500 MHz, it will outperform
a P100 too.. All this proves is that the P-100 has a superior
architecture to the 5x86 since it can provide the same level of
performance at lower clock speeds.. On the other hand if the results
had been that the 5x86 at 90 or 100 MHz ran faster than Pentium
at 133 or 160 MHz, then that would be something to talk about...
--
Christopher Lau | Nortel Wireless Networks / Bell-Northern Research Ltd.
Mr. Unix | This article contains my own opinions, not BNR's.
-- | Bring back Trudeau!
cc...@bnr.ca | (613)-763-8392
><PC config deleted>
I believe your assumption would be valid if they were using the same
bus but they don't. The Cyrix and AMD 5x86 chips are using a 486
32-bit bus NOT a Pentium 64-bit bus to get near/better pentium speeds
at higher internal/external clock rates. After all P90/100 has
external bus speeds of 60/66 respectively while an AMD 5x86 133 has a
bus speed of 33 and a Cyrix 5x86 120 has bus speed of 40 MHZ. Now what
I would like to see is a Cyrix 5x86 120 with a 60MHZ BUS go at it with
a P90 and I do believe if that were possible the Cyrix chip would kick
the P90's ASS.
For $450 you could have purchased a P5-133 with a Triton MB from Fry's
electronics.
--Joel Sartoris
>
> The point is that these 5x86 chips are performing at P90/P100 levels
> at a fraction of P90/P100 costs.
>
Until you've replaced the overclocked (burned-out)cpu a few times. I guess to
be fair to the P90/100s, they should be overclocked and these (however
inadequate) benchmarks rerun.
--Joel Sartoris
You can't. You may upgrade to the intel DX4/100 overdrive, which runs at 5v.
Paco
That's just blatently WRONG.
You *CAN* get a voltage regulator for the AM586 133mhz chips (or for any
3.3v chip for that matter) to use it in a 5v motherboard. They aren't WIDELY
available, but you can get them. They cost around $40.
- Turbo
[tu...@trend1.com]
The correct answer is: It depends.
The PCI bus runs at the speed you set the PCI bus clock to. The PCI bus
clock can be a derivative of the mainboard clock or a separate clock.
<Insert lengthy discussion of buffers here.>
With a separate clock, you could make it run at a constant 33MHz, reguardless
of what speed the system clock is running.
When deriving the clock from a single mainboard clock, the PCI will
typically be clocked at 25, 30 or 33 MHz.
--
Rick
>Is anyone running the AMD CPUSPEED.EXE program to benchmark their overclocked CPU?
>And if so what are you getting?
>CPUSPEED is a benchmark program found on the AMD homepage.
>Just wondering..
>Bryan Frain
Where is this program located on the AMD Site?
I looked all thru the Support section and other likely areas with no
luck.
Thanks.
Greg
>Yousuf Khan (yk...@achilles.net) wrote:
>: With your current DX2-66, you can successfully overclock it to 80Mhz.
>: I've done just that with an AMD DX2-66.
>A bit too optimistic, according to a survey, the AMD DX2-66 only appears
>to successfully overclock roughly 70% of the time. Problems can be subtle
>and difficult to notice.
Been running it for a year and a half now. Of course, I got the CPU
fan, maybe one should take that as a proviso, but when I bought the
motherboard as a 66Mhz system, they insisted on putting that fan on
anyways, despite the fact that the chip was supposed to run at 66Mhz
from the factory.
Yousuf Khan
The rationale is the economics.
$85 for the 5x-160
$300 for the P100 + $200 for a board + another 150-200 for a PCI video card.
Mark
My well spent 2 cents worth.
>Is anyone running the AMD CPUSPEED.EXE program to benchmark their overclocked CPU?
>And if so what are you getting?
>CPUSPEED is a benchmark program found on the AMD homepage.
Reports my Cyrix 5x86-120 as a 150MHz CPU.
--
Nigel Barker
ni...@openvms.demon.co.uk
It's different in the PC world: (INTEL, AMD, CYRIX)
Each new generation in the X86 cost more. The chips use the same
logic (CMOS) but technology keeps allowing designers to put more
gates in a chip and run it faster. That's why they come out with
newer improved chips. AMD and CYRIX have used new technologies to
improve on the the 32 bit 486 CPU. INTEL focuses their attention on the 64 bit
586 CPU. INTEL charges premium prices for this new technology.
AMD and CYRIX charges around $90 bucks for new technology in a 32 bit CPU.
If a 486 32-bit processor runs DOS and WINDOWS as fast as a 586 processor
and only cost $90 bucks... I don't think I'll buy that 586 (I'd have
to buy a motherboard and video card to go with it to).
be cool,
Wayne
>>: Sysinfo 8.0 PC Config 8.03 3d-Bench
>>: Cx5x86-120 ~260-380 (?) 19250% 90.9
>>: Pentium-90 288.3 19850% 83.3
>>: Pentium-100 316.9 24350% 100.0
>>: AMD 5x86-133 288.7 18050% 90.9
>>: AMD 5x86-160 346.1 23550% 100.0
>>: Pentium-120 380.2 28850% 111.1
>>: Pentium scores (90, 100 MHz) taken in:
>>: Intel Premeire II MB, Neptune chipset, 256k cache. Diamond Stealth 64
>>: Pentium 120 score:
>>: Taken in my other PC w/ Triton chipset, 512k PB cache, Hercules
>>: Stingray/64
>>: Cyrix 5x86 scores:
>>: Averages taken from various sources
Hello to you all in this group.
I've been of the net for about a year, and trying to fit in again.
Does anyone have an adress where I can obtain the benchmarking
software ? It will be nice to do some posting again.
I have a P-120 (Intel), and it overclocks fine.
Greetings,
Jan
As I speak;
ing. Jan Pieter Bolhuis, Professional Software Engineer.
Currently working for the Royal Netherlands Navy as a
Weapons Engineer and Teacher of Technical Computer Sciences.
I can be reached on (+31)(0)20-520-2466, or on the net
bol...@xs4all.nl
:-))
>
> Also, to add on, Since you have a PCI bus, your speed will be 20mhz for the
>when you run the chip at 160mhz as opposed to 33mhz at 133mhz..I believe
>the current limit for your PCI is 33mhz...but that will soon change.
> Either way I have mine at 160mhz and see no reason why I would want
>to run it at a slower speed just to speed the video bus speed. It
>runs everything flawlessly and never chops!
Dosn't the PCI bus stick to 33 MHz no matter what the processor clock
is set to?
I thougth this was one of the PCI bus advantages over Vesa?
**********************************
* *
* Zax last of the Zaxanoids *
* *
**********************************
In <4h4and$m...@rs10.tcs.tulane.edu> Bryan Frain
<bhf...@mailhost.tcs.tulane.edu> writes:
>
>Is anyone running the AMD CPUSPEED.EXE program to benchmark their
overclocked CPU?
>And if so what are you getting?
>CPUSPEED is a benchmark program found on the AMD homepage.
>
>Just wondering..
>Bryan Frain
: The options are really 100, 120, 133, 150, 160, and 200 :O. The chip can
: either run in 3x or 4x mode. So:
: Clock Multiplier Chip Speed
: -------------------------------------
: 33MHz 3x 100MHz
: 33MHz 4x 133MHz
: 40MHz 3x 120MHz
: 40MHz 4x 160MHz
: 50MHz 3x 150MHz
: 50MHz 4x 200MHz
: I have not yet run mine at 150, just at 160. So I can't say yet whether the
: additional bus speed will make up for the lower chip speed. However, tonight
: I will be trying out 150 and 200 and will post benchmarks here with results.
: Brett
: bre...@io.com
Also, to add on, Since you have a PCI bus, your speed will be 20mhz for the
when you run the chip at 160mhz as opposed to 33mhz at 133mhz..I believe
the current limit for your PCI is 33mhz...but that will soon change.
Either way I have mine at 160mhz and see no reason why I would want
to run it at a slower speed just to speed the video bus speed. It
runs everything flawlessly and never chops!
Ryan
Duane Saylor
-------------------------------
Jim:
Vic20!!! Don't be cheap! Get a C-64!!!
(I had both a C-64 and a C-128 --- boy, were those the days!!!)
Frank O'Neill, Independent Consultant
New York, NY
care to post the url or phone number??
==========>
Jim White
South Shore Software
jimw...@cent.com
www.cent.com/paradise.net
jimw...@msn.com
>mass...@risky.ecs.umass.edu (RYAN J MASSIMINO) wrote:
>
>
>>
>> Also, to add on, Since you have a PCI bus, your speed will be 20mhz for the
>>when you run the chip at 160mhz as opposed to 33mhz at 133mhz..I believe
>>the current limit for your PCI is 33mhz...but that will soon change.
>> Either way I have mine at 160mhz and see no reason why I would want
>>to run it at a slower speed just to speed the video bus speed. It
>>runs everything flawlessly and never chops!
>
>Dosn't the PCI bus stick to 33 MHz no matter what the processor clock
>is set to?
>
>I thougth this was one of the PCI bus advantages over Vesa?
depends on the MB architecture.... if PCI clock is a seperate,
asynchronous clock, it could be 33 mhz for every case... but dividing
down is easier/cheaper and asynchronous is not neccessarily very
desirable....
On 5 Mar 1996, M. Edward Rode wrote:
>
> I am deciding weather or not to buy a Quad speed changer CD Rom
> drive or a single Six speed CD Rom drive. Could someone who has used
> both quad and six speed tell me if there is any significant and or
> practical difference.
>
>
>On 29 Feb 1996 13:46:21 GMT, Bryan Frain
><bhf...@mailhost.tcs.tulane.edu> wrote:
>>Is anyone running the AMD CPUSPEED.EXE program to benchmark their overclocked CPU?
>>And if so what are you getting?
>>CPUSPEED is a benchmark program found on the AMD homepage.
>Reports my Cyrix 5x86-120 as a 150MHz CPU.
>--
>Nigel Barker
>ni...@openvms.demon.co.uk
The AMD cpuspeed program reports my AMD586-133 as either
a 120 or 150 depending on how many times I run it. It never
reported 133, go figure.
Running this same chip at 160, the program reported 150.
Seems accurate for speeds of 120 and below.
-Chris
>mass...@risky.ecs.umass.edu (RYAN J MASSIMINO) wrote:
>>
>> Also, to add on, Since you have a PCI bus, your speed will be 20mhz for the
>>when you run the chip at 160mhz as opposed to 33mhz at 133mhz..I believe
>>the current limit for your PCI is 33mhz...but that will soon change.
>> Either way I have mine at 160mhz and see no reason why I would want
>>to run it at a slower speed just to speed the video bus speed. It
>>runs everything flawlessly and never chops!
>Dosn't the PCI bus stick to 33 MHz no matter what the processor clock
>is set to?
>I thougth this was one of the PCI bus advantages over Vesa?
>**********************************
>* *
>* Zax last of the Zaxanoids *
>* *
>**********************************
No, The PCI has a maximum of 33 MHZ, soon to be 150, but oh well. As
for the Dillitante who raised his processor to 160MHz well good luck..
Can you actually notice a difference in speed?!! Well if you can I
will be very surprised, considering that you still are running your
33mhz bandwith external you just kicked up your internal processing to
160.. You are asking for Probs! .. O well just my Silver Dollars
Worth HEHE.. The only way you would maintain a 20mhz external would
be if you moved a jumper on the board to be 20x8 Which Would be
STUPID, since you could easily take a 33x8 which would be much faster
oh well.. Your video MUST SUCK!!!!! Do a sPecINt, and a pcBench,
and a winbench,, bet you mine beats yours down...
Ursius
We have excellent experience using the Goldstar 4X EIDE drives. As for the
debate 4x vs. 6x vs. ?x, my understanding is that no software is wriiten to take
advantage of DTR beyond 4x, and the only possible senerio that utilizes the
speed beyond 4x would be if you had a back-up burned onto a cd, then were
restoring all your data from that CD. Software installation generally is not
DTR dependant, although I'm sure some will chime in that "unknown OS that
install faster by x factor".
Plain truth is no standards exist beyond 4x, and I've not heard of any MultiMedia
software that is written for anything other than 4x, lots is still written for 2x
cd-rom speed.
This year will bring tons of Optical drives at low prices. These will quickly
replace cd-rom drives as the opticals will be backwards compatable to cd-rom
formats for read, but will be read-rewrittable with the proper disks. These new
drives should be under $500 someday soon...
Rick Lindsay
Lindsay Computer Systems
Voice: 512-719-5257
Fax: Available by request...
Hey dumbass go look at this page and tell us all what you
see. http://www.amd.com/html/overview/corppr/9619.html
Well I'll tell you anyways. THESE TWO COMPANIES ARE THE SAME!!!!!
lamer
I have done the same (overclocking the AMD5x86-P75-133Mhz)
and I would just like to now what kind of combination of heatsink and
fan are most suiteable (read most safe!) for this situation ?
Is it better to choose Write throuh or Write back regarding the
internal L1 cache ?
Otherwise - this little baby is just running fine (so far...)
Kenneth
You're mistaken. On a VL/PCI system the PCI bus hangs off of the VL-BUS,
not vice-versa. The VL bus *will* clock to 40MHz (or 50 even) and it
takes the PCI up along with it. My 160MHz AMD 5x83 is running on a 40MHz
bus -- memory, VL, and PCI. Even if PCI didn't clock up with the rest of
the system and stayed at 33MHz, everything else is faster -- video and
memory.
My OS/2 benchmarking program shows that I run neck-n-neck with my P90 --
slightly faster in some test, slightly slower in others. My VL-bus S3 864
shows nearly identical numbers to the PCI S3 864 on the P90 here at work.
> .. O well just my Silver Dollars
>Worth HEHE.. The only way you would maintain a 20mhz external would
>be if you moved a jumper on the board to be 20x8 Which Would be
>STUPID, since you could easily take a 33x8 which would be much faster
>oh well..
I have no clue what you're spouting on about here -- and I *know* what I'm
talking about...
>Your video MUST SUCK!!!!!
Like I said, it benchmarks nearly identically to my P90 with the same
video chipset. My P90 (PCI s3 864) gets 54.270 PM-Marks (OS/2 benchmark).
My 486/160 (VL s3 864) get 53.450 PM-Marks.
> Ursius
Get a life.
>Hey dumbass go look at this page and tell us all what you
>see. http://www.amd.com/html/overview/corppr/9619.html
> Well I'll tell you anyways. THESE TWO COMPANIES ARE THE SAME!!!!!
>lamer
Well, Lamer(did your mom name you that?)
You're entitled to your opinion, but if I'm a dumbass you could at
least tell me why or make some kind of point. Sure NexGen has merged with
AMD. If that makes me a dumbass, I guess that's something I'm going to
have to live with. At least my mom still loves me.
Have fun Lamer before your mom comes home.
Gawd. Aren't kids great?
Also, most apps can't use the six speeds capabilities. Few can
even use the quads, they revert to double speed. If you are
accessing databases and installing software from the CD ROM,
then the six speed will help, otherwise go for the quad changer.
Another thing to consider is that the CD ROM will be
dissapearing in a while. The new DVD is supposed to be out late
this year and who knows, in two years you may not even want to
have an "old" CD Rom drive in your system.
So don't spend too much.
--
Lorne
Ideas in Motion
id...@interlog.com
http://www.interlog.com/~ideas
<snip>
>>>>
> ????????;-/????????
> Man, to see the determination of some of these folks is amazing.
>
>
> For $199 you can get a NexGen VL P90, cpu, hs, mb and fan at:
>http://hug.leba.net/pc/specials.html
>
> And guess what. They even use the label "586" in their name! Correct
>me if I'm wrong here but I'm going to guess that an AMD 586 would turn to
>glass before it could even come close to the performance of a
>NexGen VL P90. And a tweaked to the edge NexGen, which would be a more
>accurate comparision (considering educated and intelligent people
>were in charge of classifying their prospective cpu's) would probably
>result in yet even more interesting results, like a small AMD supernova
>due to the fusion and then implosion from the intense heat produced by
>it's suicidal mission of pretending to be something other than what it
>was originally designed-KABOOOOM!!!.
>
>
Yeah, but can you find a NexGen with a math-co on it? Fact is,
todays apps EXPECT you to have the math co, so you ARE going to need
it. You can try to emulate it via software, but many software
packages are not too keen about that...
Yes, it is a smoking chip, but because it does not have a math-co
(the Cyrix 5x86 has a spectactular one), it is useless for most
people.
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
Chris Swan---->>chr...@umr.edu
"You see? I share it all with you, even the greatest mystery
of all time, the mystery by which I compose my life. I reveal
it to you in words:
"The only past which endures lies wordlessly within you."
God Emperor of Dune--Frank Herbert
-------------------
//////////////////////////////
Do you happen to have any measurements of the Nx586-fp's FPU performance?
At the moment the Pentium and the Pentium Pro have no rivals in the
x86 market for FPU performance. (The Cyrix's bench at 50-60% of
the equivalent Pentium.)
--
C. Chan <ch...@alfrothul.uchicago.edu> | .chigless in Chicago
Is there a pattern here?
Will <w-de...@ux7.cso.uiuc.edu> wrote:
>Strange, AMD's utility reported my Cyrix 5x86/100 in a Shuttle 433 as a
>150Mhz 486.
> Will DeHaan ~ tec...@uiuc.edu
>On Fri, 1 Mar 1996, Nigel Barker wrote:
>> >Is anyone running the AMD CPUSPEED.EXE program to benchmark their overclocked CPU?
>> >And if so what are you getting?
>> >CPUSPEED is a benchmark program found on the AMD homepage.
>>
>> Reports my Cyrix 5x86-120 as a 150MHz CPU.
>>
>> --
>> Nigel Barker
>> ni...@openvms.demon.co.uk
>>
>>
"Eric Witlin's Law Page"
http://www.netaxis.com/~ehwitlin
My experience running the AMD CPUSPEED program is that it cannot
reliably measure anything above 120Mhz. Try a different program. Even Norton
SYSINFO seems better.
As an email correspondent of mine said "I sure hope the AMD CPU chips
are more reliable than their CPUSPEED program..."
These were in Bytemarks, which had been actually penalizing the
Pentium Pro because of Watcom and MS C malloc alignment. The
SPEC suite also shows that the Pentium Pro has dramatically
beefed up FPU performance on a par with the better RISC CPUs.
> I have yet to see Pentium at least outpace a
>comparable Cyrix processor in FPU performance when it comes to apps
>that use it anyway.
I'm not a fan of Intel, and I use RISC chips at work and a Cyrix
CPU at home. But I don't really believe this statement. The Cyrix
6x86 chip is their Pentium class CPU, and it gives up FPU performance
to gain some advantages in some integer (it has 3 integer stages
to the Pentium's two; to maintain a reasonable die size, something
had to give.) The 5x86 series has a better FPU than an equivalently
clocked Intel or AMD 486 (but it seems a number of people have been
able to overclock their AMDs) but doesn't outperform a Pentium's.
>The Nexgen CPU at least match the Pentiums in FPU
This is what I'm asking. Until recently, NexGens did not have
an integrated FPU.
>but the latter ones surpass it because they are running at the actual
>clock speed instead of being comparable to a relative Pentium like the
>processor they make with speeds at 100MHZ or less.
>
Could you run some benchmarks and tell us how well the NexGen
NX-586-fp's FPU performs?
>In <Pine.OSF.3.91.960305...@ucs.orst.edu>, Gregory Straw <str...@ucs.orst.edu> writes:
>>M. Edward:
>>The issue is not 6x vs 4x the issue is that the drive spec.'s
>>themselves. What I mean is that you are trying to compair apples and
>>oranges. You need to look at what the buffer is (256k is
>>recommended) and what the seek times are. Then the transfer rate at a
>>prolonged period. The last thing I would say is you can find some 4x
>>CD-Roms that are VERY FAST I think that ACER, TEAC are the brands that
>>you want to check out. WHAT EVERY YOU DO SAY AWAY FROM CREATIVE LABS
>>6X..... I've seen nothing but problems and the performance vs $$ does not
>>equal.
>>
>>
>>On 5 Mar 1996, M. Edward Rode wrote:
>>
>>>
>>> I am deciding weather or not to buy a Quad speed changer CD Rom
>>> drive or a single Six speed CD Rom drive. Could someone who has used
>>> both quad and six speed tell me if there is any significant and or
>>> practical difference.
>>>
>>>
>We have excellent experience using the Goldstar 4X EIDE drives. As for the
>debate 4x vs. 6x vs. ?x, my understanding is that no software is wriiten to take
>advantage of DTR beyond 4x, and the only possible senerio that utilizes the
>speed beyond 4x would be if you had a back-up burned onto a cd, then were
>restoring all your data from that CD. Software installation generally is not
>DTR dependant, although I'm sure some will chime in that "unknown OS that
>install faster by x factor".
>Plain truth is no standards exist beyond 4x, and I've not heard of any MultiMedia
>software that is written for anything other than 4x, lots is still written for 2x
>cd-rom speed.
>This year will bring tons of Optical drives at low prices. These will quickly
>replace cd-rom drives as the opticals will be backwards compatable to cd-rom
>formats for read, but will be read-rewrittable with the proper disks. These new
>drives should be under $500 someday soon...
>Rick Lindsay
It always seems that Optical drives are just around the corner.
And that they'll replace CD-ROMs. I'll just say one thing; DVDs. By
the middle to end of 1997, they could be on its way to completely
replace CDs, CD-ROMs, and LaserDiscs. VHS, and Cassette tapes could be
surplanted shortly afterwards with a recordable device. And with the
manufacturer's behind the standard (basically, it could be easier to
list who isn't going to support it), I don't see how Opticals are
going to be anything but a flash for the general market.
William Kang
73430...@compuserve.com
What?
> The 16 bit is a bit faster, and the 32 bit is truely faster.
What 16bit or 32bit?
>Right now I can get a NexGen pf100 Mhz PCI board for only $295. It is RISC
>arch, right now Intel is trying to go RISC. The PentPro is a RISC platform
>on a CISC chip. That is the reason why it runs slower than the Pentium on 16
>bit.
What?
>Wiggie
Wiggie, please clarify. You are on another planet. (I Appreciate that english
may not be your first language also.)
DJF
PCI-SPEED OR NOT !
Since I tried this overclocking thing, I allmost burnt out
my DiamondStealth64Video card !
For about a week after I pumped up the external clock-speed to 40 Mhz
I noticed some weird "gost"-trails to the right of dialog-boxes. Later
this continued in DOS with white trails after caracters (mostly on
white text / Black screen).
Then one morning - working on the system for about an hour- the left
part of the screen started "pumping" - seconds after that the picture
disappeared/came back in pumping kind of way. Turned of the machine
VERY fast !
Opened up the case and started the computer again....
After a couple of minutes, I gently touched the 2 big chips on the
grafic-card (the RAMDAC-175Mhz and S3-968) THEY WERE VERY HOT !!!!
So I think that my PCI speed IS 40Mhz (and not 40 / 2 = 20) - I have
never had these problems before !
My solution: Placed a heatsink on each chip (+heat-paste) and a little
CPU-fan nearby - no problems since !
Anyone outthere experienced the same ?
Regards kenneth R
: These were in Bytemarks, which had been actually penalizing the
: Pentium Pro because of Watcom and MS C malloc alignment. The
: SPEC suite also shows that the Pentium Pro has dramatically
: beefed up FPU performance on a par with the better RISC CPUs.
I'm not sure but an 'old' 75mhz R8000 is rated at .3 gigaflop , how much
better is the all new 200mzh P6? dramaticly better mean ~1 gigaflop,
maybe the p6 chip alone is worth the price of a loaded computer afterall.
Stephan
Mike
: Also, most apps can't use the six speeds capabilities. Few can
: even use the quads, they revert to double speed. If you are
That's complete crap. If you have, for example, an 8x drive, the apps. will
benefit the speed. If the mpegs are not smoother, the load time will be
GREATLY shortened.
Shine On
>: These were in Bytemarks, which had been actually penalizing the
>: Pentium Pro because of Watcom and MS C malloc alignment. The
>: SPEC suite also shows that the Pentium Pro has dramatically
>: beefed up FPU performance on a par with the better RISC CPUs.
>
> I'm not sure but an 'old' 75mhz R8000 is rated at .3 gigaflop , how much
> better is the all new 200mzh P6? dramaticly better mean ~1 gigaflop,
> maybe the p6 chip alone is worth the price of a loaded computer afterall.
>
> Stephan
I usually go by the SPEC suite and the '92 numbers for now since
not all the CPUs have been benched under SPEC 95 by independent
testing labs. The R8000 isn't old, just middle-aged; we use it
in some of our Onyx and Challenge boxes. The R8000 vs P6 is an
apple and oranges comparison; the R8000 is specifically designed
for high FPU performance while the P6 is a more general purpose
chip. The R8000 scores around 120SpecInt92/350SpecFlt92; the
P6 does around 200/200. So the P6 doesn't beat the R8000 in
FPU but is available in much less expensive machines. Of course
the R10k will be out soon and is of a current generation with
the P6.
In general, MPEG-1 in software, doesn't benefit from anything much over
4x. The
limitation is not disk I/O speed. Greater than 4x will provide faster
loading of S/W from CD and better indexing for things like help files
on CD.
But if you're looking for better video, get 4x, and save your money for
DVD.
Mike Schmit
-------------------------------------------------------------------
msc...@ix.netcom.com author:
408-244-6826 Pentium Processor Programming Tools
800-765-8086 ISBN: 0-12-627230-1
-------------------------------------------------------------------
Seriously, a changer is nice ans so is 4x. I had a Pioneer DRM-604X
four years ago (and who said quad speed drives were just released last
year) and while it was fun, you pay a serious penalty under DOS/Win to
get access to all those disks (up to 67K with 6 virtual CD drives).
The software that allowed you to swap disks without using additional
drive letters never worked well enough to be useful. Plus not all
controller can handle multiCD drives under every OS. Make sure you
research this out before you buy or you will be SOL.
Good luck in getting the drive.
aur...@xs4all.nl (IJdo Dijkstra) wrote:
>pas...@nyongwa.montreal.qc.ca (Pascal Durocher) wrote:
>>That's complete crap. If you have, for example, an 8x drive, the apps. will
>>benefit the speed. If the mpegs are not smoother, the load time will be
>>GREATLY shortened.
>Mpegs use a lot of CPU power.....so does running a 8x (ATAPI)....
>IJdo Dijkstra aur...@xs4all.nl
>----------------------------------------------
> http://www.xs4all.nl/~aurora
> CS student, Freelance sysadmin
CPUmark16 80 (from PC Bench 9.0) failed for WinBench every time
CPUmark32 92.1
These numbers appear low. I do not know if they are related to my limited
memory or not. My CMOS settings are set to the fastest possible. It
without a doubt is faster than the Cyrix 486/80 I had previous the
difference being very obvious in games such as Terminal Velocity. If
someone has some ideas on how I can fine tune my system Im all ears.
Benchmark Ziff-Davis' WinBench 96 Version 1.0
Display Adapter Diamond SpeedStar Pro
Display DRAM 1 MB
Display Driver SSPRO256.drv, Jun 30 10:58:50 1993, 112736 bytes,
1.10.00.00
Display Mode 1024 X 768 8 bits/pixel
Display Refresh Rate 60 Hz (not verified)
Display VRAM 0 KB
HD Controller RAM 256KB
Hard Disk Conner 840
Hard Disk Controller No-name
OS Software Cache SmartDrive 5.00: A C- D- E- F- G- H- I- J-
OS Software Cache Size 256 KB, Windows: 256 KB
Off-Chip Processor Cache 256 KB
Operating System MS-DOS 6.20
Processor AMD 587 133/oc'd to 166 Family 4 Model 14 Step 4
Features 1h
Processor RAM 8 MB
Processor Speed 133 OC'ed to 160
Windows Software Cache Size 1024 KB
Windows System Windows for Workgroups 3.11 Enhanced Mode with
Paging
Mike
Wow, that is a very low CPUmark! it
should be more than twice that figure.
You have the MTI R407e, right?
many users in comp.sys.....chips
have been complaining about its poor
performance. But FWIW, you may want to
tweak your BIOS settings, settings WS's
as low as possible and DRAM readings
to fastest.
email me a copy of your CMOS settings I've seen these proeblems many times
before. Sometimes it's easy, others it's a KNIGHTMARE!
hope it's not the later.....
cheers Gordon....
Spent the evening tweaking the bios setting using CTCHIPZ. I was able to get
it up to 81.62 (CPUmark16) adjusting for fastest DRAM and fewest cache read
and write states. Even chopped the 16 and 8bit I/O cycles to the minimum.
Right now I'm convinced that my chipset (SIS471) is the culprit. I tried
running at 200mhz - no go period. I did reset to 133mhz with an appropriate
drop in performance (67 on the CPUmark16). I worked with the Write-Back and
Write-Thru options with little effect. I am running the latest AMI bios
(version F) for this board. Am considering swapping out for a Cyrix to see if
the architecture will make the difference.
Mike
Something is wrong here. It sounds like your old Cyrix 80/40 was running
too slow. On my motherboard my benchmarks are as follows:
PCBench 9.0 486DX/2-80 Am5x86-140 Am5x86-150
Dosmark: 640 840 1074
CPUMark: 103 137 167
Video: 6251 6247 6531
Disk: 469 632 907
I have a SiS-471 baseed motherboard and I have found that mose SiS-471
motherboards are very good performers. It sounds to me that your
motherboard is poorly designed or defective. What brand is it?
Is there a possibility that its under warranty? The reason
I say this is that your Cyrix 80 Mhz numbers are low as well.
For an 80 Mhz 486, your CPUMark16 from Winbench should be on the
order of 100, the 5x86 o.c. to 150 Mhz should be on the order of 150.
My motherboard is by no means perfect. It won't boot with the 5x86
in 4x mode and if I enable the internal write back cache, I loose
my L2 cache. I really don't think a Cyrix chip will solve your problem.
Just another thought, do you have a L2 cache? No L2 cache can really
slow down a system. Also, can you tell me where to get CTCHIPZ?
Hope this helps
-Mike
This is by no way 100% truth. I'll just relay what 2 different vendors told
me. I was debating between the OPTI chipset and the SiS chipset and one
vendor told me that the SiS was only capable to handle dx4/120 while the OPTI
was compatible with the 5x86. I called around and most other vendors told me
that their was no difference. Finally I stumbled around one that said the
same thing that the previous vendor told me... SiS doesn't support 5x86.
Anyone else heard anything similar?
____________________________________________________________________
Chris Richards | rich...@tamu.edu | http://grover.bre.com/~richards
Thier chipset manual specifically states that the SiS 471 supports the
following processors: 80486DX2/DX/SX/SL Enhanced, *P24D*/P24T/P24C and
M6/M7. Note the *P24D*. AMD claims the Am5x86 is P24D compatible.
All P24D means is a 486 processor with internal write back cache -
sounds like the Am5x86 to me!
On my SiS-471 MB, I cannot run in 4x mode nor can I enable the internal
WB cache, 4x won't boot, and internal WB cache disables my external
cache. It appears to be an Award BIOS problem. For instance,
the Award bios shuts off support for P24D write back cache, even
though it is a "don't care" for 486. I did quite a bit of research
when I bought the mother board and I found the the SiS-471 boards
were low cost, very high performance, and very stable. I have little
reason to doubt AMD or SiS on this one. My biggest problem seems to
be Genoa (board manufacturer) and Award.
Of course, if you are talking about the Cyrix 5x86 chip, then the SiS
documentation does not mention it. I think this fellow has 1) a bad
(defective) mother board, 2) really screwed up CMOS (not likely, al-
though maybe disconnecting the battery to clear the CMOS would fix
it, again not likely, but worth a try) or 3) a poorly designed mother-
board.
-Mike
>Follow up
>CPUmark16 80 (from PC Bench 9.0) failed for WinBench every time
>CPUmark32 92.1
>These numbers appear low. I do not know if they are related to my limited
>memory or not. My CMOS settings are set to the fastest possible. It
>without a doubt is faster than the Cyrix 486/80 I had previous the
>difference being very obvious in games such as Terminal Velocity. If
>someone has some ideas on how I can fine tune my system Im all ears.
>Benchmark Ziff-Davis' WinBench 96 Version 1.0
>Display Adapter Diamond SpeedStar Pro
>Display DRAM 1 MB
>Display Driver SSPRO256.drv, Jun 30 10:58:50 1993, 112736 bytes,
> 1.10.00.00
>Display Mode 1024 X 768 8 bits/pixel
>Display Refresh Rate 60 Hz (not verified)
>Display VRAM 0 KB
>HD Controller RAM 256KB
>Hard Disk Conner 840
>Hard Disk Controller No-name
>OS Software Cache SmartDrive 5.00: A C- D- E- F- G- H- I- J-
>OS Software Cache Size 256 KB, Windows: 256 KB
>Off-Chip Processor Cache 256 KB
>Operating System MS-DOS 6.20
>Processor AMD 587 133/oc'd to 166 Family 4 Model 14 Step 4
> Features 1h
>Processor RAM 8 MB
>Processor Speed 133 OC'ed to 160
>Windows Software Cache Size 1024 KB
>Windows System Windows for Workgroups 3.11 Enhanced Mode with
> Paging
>Mike
Hey Mike those Numbers are very low. I know were talking apples and
oranges here but a recent run on Winbench 96 ver.1.0 on my P54C-120Mhz
yielded scores of 205 for CPUmark16 and 204 for CPUMark 32. I thought
these numbers were low but I don't have the benefit of Pipeline burst
SRAM or EDO RAM. Your numbers are about the performance of a 100Mhz
DX4 or at best a P54C-66. What were your expectations at clocking up
to 160Mhz? I am interested in performance of the newer
Chips/Motherboard Combos.
[snip]
>I found the the SiS-471 boards
>were low cost, very high performance, and very stable. I have little
>reason to doubt AMD or SiS on this one. My biggest problem seems to
>be Genoa (board manufacturer) and Award.
>
Do you have any opinions on the OPTI chipset? I had the choice b/w the OPTI
and SiS on a QDI motherboard and I went with OPTI because of what the two
vendors had told me....
You are getting a 50% discount here.......
: Even chopped the 16 and 8bit I/O cycles to the minimum.
: Right now I'm convinced that my chipset (SIS471) is the culprit.
Probably not, I've heard of others with the 471 doing fine.
The board can be the problem though.
--
~Frisky Kitty~
!*% T-3..............T-2.............T-1.............BOOM.....................
-Mike
What a bunch of bulls***. SiS 471 and SiS 496/7
all run fine with both Cyrix and AMD 5x86 chips. If a given
motherboard can't handle a certain CPU, the problem lies elsewhere...
not the chipset....
Also, SiS and Intel chipsets are by and far the best on the market;
The Intel Triton and SiS 501/2/3 both have similar CPU performance
however the Triton has slightly better PCI performance.