Article View: comp.os.ms-windows.win95.setup
Article #623Re: Cyrirx6x86-166P+: The Fastest PC for DOS/Windows World
From: Luka
Date: Wed, 03 Apr 1996 17:06
Date: Wed, 03 Apr 1996 17:06
281 lines
18722 bytes
18722 bytes
Dana H. Taylor Jr. wrote:
>
> >It's okay, Intel has the P55C coming out by the end of the year, with
> >the new MMX instructions, as well as larger Icache and Dcache, so the
> >performance per Mhz will improve as well.
>
> I just bought a Cyrix P120 (100MHZ) 686. I have it over-clocked to 120
> MHZ. (20 MHZ greater than what it should be at) It goes super fast! I
> only paid $209 for the CPU and $189 for an ASUS PCI motherboard. I am
> very happy with the performance so far. It goes a lot faster than my
> Pentium 100 at work. Any way, my question is : Is it dangerous to
> over-clock the CPU? Should I set it back to 100. Will my system
> over-heat?
>
> Any help would be appreciated,
> Dana
I found this post in one of the newsgroups a while ago and I saved it.
It gives you statistics on overclocking (not written by me; the guy's
sig is at the end)
----------------------------------------------------------------------
Subject: Re: Overclocking is safe(almost)..<Read this>
Date: Sat, 09 Mar 1996 00:47:53 +0100
From: Stian Larsen <stlarsen@telepost.no>
Organization: TelePost Public Access
Newsgroups: alt.games.quake,rec.games.computer.quake.misc
References: <4gbtsc$j5p@sundog.tiac.net> <4gdpt3$qhh@library.erc.clarkson.edu>
<312e5088.492330@news.neosoft.com> <4hd21q$8nl@sdcc12.ucsd.edu> <4hhk1i$f9r@jari
Here's a doc i got:
Many individuals have overclocked systems and they have experienced no
recognizable problems. CPU manufacturers warn that problems may not result
in an immediate catastrophic failure. That there may only be very subtle
and difficult to notice problems. For example, only one instruction with a
particular data pattern may fail. Additionally, they warn that there may
be long term cumulative damage that reduces the lifespan of the CPU. This
survey was an attempt to gather some data on the reliability, or at least
perceived reliability, of overclocking. It also attempts to gather data on
how many users are willing to overclock important systems. This survey is
not an endorsement of overclocking, proceed at you own risk (physical
damage, run-time reliability).
This survey is over. The questions and results are presented together in
this document to help you interpret the results. Please do not respond
with additional results. Other questions or comments may be sent via
e-mail to:
adtribelli@acm.org
-------------------------------------------------------------------------
What type of CPU do you have?
Manufacturer:
Model:
Speed (manufacturer recommended):
Cooling (none, heat sink, fan, peltier, ...):
At what speed (overclocked) was it run at?
Did it work?
If no, was the problem obvious (system hangs) or subtle (one program
misbehaves)?
Did the problem go away when the CPU was returned to its recommended
speed?
If no, why do you think this was an overclocking problem then?
If the machine being tested experiences a subtle failure (generates bad
data) in the future, is it possible that there will be a significant
amount of time or money lost (Yes/No)?
Any interesting observations?
Any suggestions on how to make a system work?
-------------------------------------------------------------------------
Results so far (15Jul95-24Jan96):
Failure Important Pass
CPU MHz Cooling Tested Passed Obvious Subtle Systems Rate
------------------- --- ------- ------ ------ ------- ------ --------- ----
Harris 286-16 20 none 1 1 0 0 1 100%
Intel 386DX-16 20 none 1 1 0 0 0 100%
25 none 1 1 0 0 1 100%
AMD 386SX-25 33 none 1 1 0 0 0 100%
AMD 386DX-40 50 none 1 0 1 0 0 0%
sink 1 0 1 0 1 0%
Intel 486SX-20 25 fan 1 1 0 0 0 100%
33 none 1 1 0 0 0 100%
Intel 486DX-20 33 none 1 1 0 0 0 100%
Intel 486SX-20 33 fan 2 2 0 0 0 100%
Intel 486SX-25 33 none 4 4 0 0 2 100%
sink 1 1 0 0 1 100%
fan 10 10 0 0 2 100%
40 none 2 2 0 0 0 100%
sink 2 2 0 0 1 100%
fan 8 7 0 1 4 88%
50 fan 3 3 0 0 0 100%
487SX-25 33 fan 1 1 0 0 1 100%
40 fan 1 0 1 0 1 0%
AMD 486SX-25 40 fan 1 1 0 0 1 100%
Intel 486DX-25 33 fan 1 1 0 0 1 100%
Intel 486SX-33 40 none 5 4 1 0 4 80%
sink 1 1 0 0 0 100%
fan 9 6 2 1 3 67%
50 none 2 1 1 0 0 50%
fan 4 2 2 0 1 50%
AMD 486SX-33 45 none 1 1 0 0 0 100%
UMC U5S-33 50 fan 1 1 0 0 0 100%
Intel 486DX-33 40 none 5 2 3 0 3 40%
sink 3 2 1 0 2 67%
fan 11 11 0 0 8 100%
50 sink 2 2 0 0 1 100%
fan 1 1 0 0 0 100%
66 sink 1 0 1 0 1 0%
AMD 486DX-33 40 fan 1 1 0 0 1 100%
AMD 486DX-40 50 none 1 1 0 0 1 100%
fan 4 4 0 0 2 100%
peltier 1 0 1 0 0 0%
Cyrix 486SLC-40 50 ? 1 0 1 0 0 0%
Cyrix 486DX-40 50 sink 1 1 0 0 1 100%
UMC U5 Super-40 50 fan 1 0 0 1 1 0%
IBM BLDX2-66 80 none 1 1 0 0 0 100%
fan 1 0 1 0 1 0%
Intel 486DX2-50 60 fan 2 2 0 0 2 100%
66 none 1 1 0 0 0 100%
sink 3 3 0 0 2 100%
fan 10 8 2 0 6 80%
80 fan 3 2 1 0 2 67%
peltier 1 1 0 0 0 100%
AMD 486DX2-50 66 fan 1 1 0 0 0 100%
Cyrix 486DX2-50 66 sink 1 1 0 0 1 100%
fan 1 1 0 0 1 100%
100 * sink 1 0 1 0 0 0%
Intel 486DX2-66 80 none 4 0 4 0 2 0%
sink 5 0 5 0 2 0%
fan 50 36 11 3 28 72%
peltier 2 2 0 0 2 100%
84 fan 1 0 1 0 1 0%
100 fan 1 1 0 0 1 100%
AMD 486DX2-66 80 none 2 2 0 0 1 100%
sink 3 2 1 0 3 67%
fan 10 7 2 1 8 70%
100 fan 4 4 0 0 4 100%
Cyrix 486DX2-66 80 sink 3 1 2 0 0 33%
fan 4 2 2 0 0 50%
SGS-Thom. 486DX2-66 80 fan 2 2 0 0 1 100%
84 sink 1 1 0 0 0 100%
90 sink 1 0 1 0 0 0%
AMD 486DX2-80 100 none 1 0 0 1 1 0%
sink 1 1 0 0 0 100%
fan 20 14 5 1 11 70%
peltier 1 1 0 0 1 100%
120 fan 3 0 3 0 3 0%
peltier 1 1 0 0 1 100%
Intel 486DX4OD-100 120 sink 1 1 0 0 0 100%
fan 1 1 0 0 0 100%
Intel 486DX4-100 120 ? 1 0 1 0 0 0%
sink 2 0 1 1 0 0%
fan 10 1 8 1 6 10%
150 sink 1 0 1 0 0 0%
fan 1 0 1 0 0 0%
AMD 486DX4-100 120 none 1 1 0 0 1 100%
sink 1 0 1 0 1 0%
fan 26 14 9 3 13 54%
150 fan 1 1 0 0 0 100%
AMD 486DX4-120 130 fan 1 1 0 0 1 100%
133 none 1 0 1 0 1 0%
150 fan 4 0 4 0 2 0%
peltier 1 0 1 0 1 0%
Intel P24T-63 83 fan 1 1 0 0 0 100%
Intel P24T-83 100 fan 2 2 0 0 1 100%
Intel Pentium-60 66 sink 3 2 1 0 1 67%
fan 6 4 1 1 5 67%
peltier 1 1 0 0 1 100%
Intel Pentium-75 90 sink 6 6 0 0 1 100%
fan 46 44 1 1 24 96%
peltier 1 1 0 0 1 100%
100 sink 3 3 0 0 2 100%
fan 39 32 6 1 25 82%
peltier 2 2 0 0 1 100%
120 sink 3 2 1 0 2 67%
fan 12 7 5 0 5 58%
peltier 2 1 1 0 0 50%
133 fan 5 2 3 0 2 40%
peltier 1 1 0 0 1 100%
150 * fan 2 0 2 0 2 0%
166 fan 1 0 1 0 1 0%
180 fan 1 0 1 0 1 0%
200 fan 1 0 1 0 1 0%
Intel Pentium-90 100 ? 1 1 0 0 1 100%
sink 5 3 2 0 3 60%
fan 64 51 11 2 41 80%
peltier 6 5 0 1 2 83%
113 fan 1 0 1 0 1 0%
120 sink 4 4 0 0 3 100%
fan 14 8 6 0 9 57%
peltier 6 4 2 0 2 67%
133 sink 1 0 1 0 1 0%
fan 5 0 5 0 3 0%
peltier 2 0 2 0 1 0%
Nexgen P90 100 fan 2 2 0 0 0 100%
Intel Pentium-100 120 sink 2 1 1 0 1 50%
fan 19 12 7 0 16 63%
peltier 2 1 1 0 1 50%
133 sink 1 0 1 0 0 0%
fan 6 0 5 1 3 0%
peltier 2 1 1 0 1 50%
Cyrix 5x86-100 120 fan 3 3 0 0 2 100%
Intel Pentium-120 133 sink 2 2 0 0 1 100%
fan 11 9 1 1 6 82%
peltier 2 0 0 2 1 0%
Intel Pentium-133 150 fan 5 2 2 1 3 40%
166 fan 2 1 1 0 2 50%
180 fan 6 4 2 0 2 67%
Motorola PPC601-66 80 fan 1 1 0 0 1 100%
OVERALL 613 426 162 25 341
69% 26% 4% 56%
* Marks experiments where permanent damage was observed and reported:
Cyrix 486DX2-50 => 100, would not boot when returned to 50Mhz.
Intel Pentium-75 => 150, burning/smoke after 5 minutes of use.
Typical Problems reported by those surveyed:
% of
failed
systems
-------
CPU permanently damaged: 2 1.07%
Would not boot: 74 39.57%
Hung or reset soon after boot: 46 24.60%
Ocassional hanging, resets, crashes, or odd behaviour: 58 31.02%
Ran fine at first but developed problems at a later date: 4 2.14%
Successful corrective actions reported by those surveyed:
% of
passed
systems
-------
Had to adjust wait states for memory: 17 3.99%
Had to disable the CPU's internal cache: 11 2.58%
Had to disable the CPU's external cache: 7 1.64%
Had to disable MB mode-3 and 32-bit transfer to boot HD: 1 0.23%
Had to adjust BUS speed: 14 3.29%
Had to use heat sink compound: 2 0.47%
Had to add CPU fan: 5 1.17%
Had to add peltier cooler: 2 0.47%
Had to use an additional system fan: 3 0.70%
Had to use an additional CPU fan card adjacent to the CPU: 2 0.47%
Had to cool CPU and PC chipset: 1 0.23%
Had to open system case: 8 1.88%
Had to use air conditioner to keep room below 85 deg F: 1 0.23%
Had to change voltage jumper from 3.3v to 3.5v: 1 0.23%
Interesting observations reported by those surveyed:
An early stepping of the CPU failed, a later stepping seemed to work.
A decrease in performance of an accelerated PCI video card.
An increase in performance of an EIDE disk transfer rate.
CPU works, but graphics card fails.
CPU works, but built-in EIDE controller fails.
CPU works, but system bus runs at slower speed.
Speed increase hard to notice without faster system RAM.
Extra wait states negated additional CPU speed.
Suspects CPU may have been falsely "remarked" at a higher speed.
S3 and Cirrus video adapters worked, but Tseng failed in same system.
BIOS misidentified I/O ports.
Seemed to run fine for 3 months, then random hangs were experienced.
BIOS only detected half of installed memory.
--
------------------
Tony Tribelli
adtribelli@acm.org
-------------------------------------------------------------------------
Decide for yourself (I decided not to overclock
Luka
Technical Details
Message-ID:
<3162B027.7330@guest.arnes.si>Path:
arc1-ca.pugleaf.net!cmir.arnes.si!usenetReferences:
<Doy952.MJA@granite.mv.net> <4jeduo$1u2@cville-srv.wam.umd.edu> <4jt7ka$bqc@news.ios.com>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!cmir.arnes.si!usenet From: Luka <luka.batistic@guest.arnes.si> Newsgroups: comp.os.ms-windows.win95.setup,comp.object,comp.os.linux.misc,comp.os.linux.networking,comp.os.ms-windows.programmer.misc,comp.os.ms-windows.win95.misc,comp.os.ms-windows.win95.setup,comp.os.os2.advocacy,comp.security.misc,comp.security.unix,comp.software-eng,comp.sys.amiga.advocacy,comp.sys.hp.hpux,comp.sys.ibm.pc.games.action,comp.sys.ibm.pc.games.strategic,comp.sys.ibm.pc.hardware.chips,comp.sys.ibm.pc.hardware.misc,comp.sys.intel,comp.sys.laptops,comp.sys.mac.system Subject: Re: Cyrirx6x86-166P+: The Fastest PC for DOS/Windows World Date: Wed, 03 Apr 1996 18:06:47 +0100 Organization: Academic and Research Network of Slovenia Message-ID: <3162B027.7330@guest.arnes.si> References: <Doy952.MJA@granite.mv.net> <4jeduo$1u2@cville-srv.wam.umd.edu> <4jt7ka$bqc@news.ios.com> NNTP-Posting-Host: ppp5-ce2.arnes.si Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 2.0 (Win95; I) To: "Dana H. Taylor Jr." <danat@tribeca.ios.com>