Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Overclocking and chess.

Author: Aaron Gordon

Date: 17:19:04 09/26/04

Go up one level in this thread


On September 26, 2004 at 12:35:03, Derek Paquette wrote:

>On September 26, 2004 at 09:33:15, Aaron Gordon wrote:
>
>>On September 26, 2004 at 00:38:17, Joshua Shriver wrote:
>>
>>>Saw an article today where a 3.6ghz machine was overclocked to 6ghz using liquid
>>>nitrogen.
>>>
>>>http://www.xtremesystems.org/forums/showthread.php?t=42655&page=1&pp=25
>>>
>>>With respect to CCC, has anyone here tried something similiar to boast game
>>>play?
>>>If it remains stable, can you imagine running on a dual 6ghz machine :)
>>>
>>>
>>>Sincerely,
>>>Joshua Shriver
>>
>>Would be fast, but I'd MUCH rather have an Ln2 cooled Athlon FX @ 4GHz. :)
>
>
>rumour has it that chips overclocked that fast, or even remotely similiar dont
>last for more than a few months.  but that is only a rumour.

The coolant systems only last a few minutes. The Ln2 isn't a maintained system.
They have a metal "cup" on top of the CPU basically. They pour the Ln2 in, run
some benchmarks while it is still cool. Super cooled CPUs do last a long time,
though. I had an old Celeron 566MHz cooled with a 172 watt peltier, cpu ran
around negative 20 degrees celsius at 1.2GHz. It ran that way for over a year
until I got my Athlon 1.0GHz (@ 1.7:). Anyway, it still runs fine to this day.
That was even running 0.08v from Intels whitepaper max vcore (2.13v, I ran
2.05v). This is just 1 of my super-cooled processors. I had old K6 chips, many,
MANY different celerons (300a @ 644, 366 @ 735, etc). All working fine to this
day.

The cooler the chip runs, the longer it'll last. This is the main problem during
overclocking. Hotter cpus means electromigration occurs faster. It is always
occuring, heat just accelerates it. Dropping the temp a LOT will allow the chip
to run much, much longer as I mentioned.

Two main things to keep your chip alive. Keep the voltage and heat down. Keep it
under the absolute maximum. If I remember correctly, old K6 (0.35um) could do up
to 3.6v w/o problems. 0.25 micron were stock at 2.2/2.4v, max of 2.8v. Current
Athlons (Tbred-A/B/Barton(0.13um)): 2.05v. As for Opterons/Athlon 64/FX chips..
1.85v.



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.