On Sunday 10 April 2005 00:26, steve elias wrote:

> additionally
> it's possible that running it without a case on the computer would do the
> same. 

Hmm, modern case designs have well-designed airflow, so much so that removing 
a case panel _might_ result in worse cooling.

Did you try blowing pre-cooled air from a portable air conditioner, or even 
from a stage fog machine, into the air inlet?

Remember Sinclair ZX80 - wouldn't work reliably without assisted cooling - a 
typical dodge was to place a refrigerated carton of milk on the housing?

> in the past when a new toshiba laptop fan failed (triggered by
> prime95 forcing it to run all the time), i was able to continue prime
> hunting by placing the laptop in our absurdly cold garage.  running it
> indoors resulted in it automatically shutting down due to over-temperature
> condition.

Why not fix the fan? Was probably under warranty anyway - the last time I 
bought a Toshiba laptop the 3 year warranty didn't say anything about failure 
due to continuous heavy use being excluded.

OTOH if it was like the Tosh 4070 I'm referring to, the fan was 
_excruciatingly_ loud and cycled on and off at regular intervals even without 
loading from CPU-intensive programs. Sometimes even if the speed was set to 
"slow" (half clock speed) and the cooling to "quiet" (rather than 
"performance"). It's spent the vast majority of the 5 years or so I've had it 
running Prime95 (well, mprime actually) in a coolish but not frigid spare 
room where the noise from the fan isn't a nuisance. And BTW the fan is still 
in full working order, despite setting full clock speed and "performance" 
cooling.
>
> on many PCs, the CPU fan needn't run while the PC is idling.

The power saving is totally insignificant. The noise reduction _may_ be 
significant - though the last few systems I've built, you really can't hear 
the CPU fan; the noisiest components are the PSU fan (~24 dBA) and the HDD 
(~22 dBA)

Another point here, if you set up the BIOS for "quiet" operation, you _may_ 
find that you have a reduced CPU clock speed most of the time - the CPU will 
throttle when the temp starts to rise, and only turn on the fan if the temp 
keeps rising.

> but with prime95 operating, the cpu fan
> has to run often on my AMD 2.3Ghz PC - 50% duty cycle, on for 2 minutes,
> off for 3 minutes.

If you can hear it, that would be irritating. Possibly more irritating than 
having it running all the time.

> on my intel P4 PCs the cpu fan will be maxxed if prime95 is running,
> 100% duty cycle...

In the case of my systems, deliberately on all the time but never at "full 
speed" because I use Zalman CPU coolers which have a speed controller in the 
fan power lead. The fan is inaudible (way below 20 dBA) if set at 1500 rpm 
and still does an adequate cooling job - the CPU temperature running Prime95 
"flat out" is around 50C (30C above ambient, and around 20C below the onset 
of thermal throttling) on my P4-3200E system.
>
> regarding whether prime95 can cause a partition failure, yes,
> of course it can.  possible mechanisms for this have been explained by
> others.  how does it go, a butterfly's wings can cause a typhoon halfway
> around the globe...

NO. Yes we have mentioned "possible mechanisms" but these are not 
specifically related to Prime95. Any program which drives the system hard 
enough will do when the effect depends on a malfunction. "Causing a partition 
failure" would depend on a direct link between code in the program and the 
effect caused e.g. some logic which, if executed on a system behaving 
according to specification, writes to cylinder 0 head 0 sector 0.

Regards
Brian Beesley
_______________________________________________
Prime mailing list
[email protected]
http://hogranch.com/mailman/listinfo/prime

Reply via email to