Mark wrote: > Agreed, David. My experience is that a machine that fails the prime95 > torture test is much more likely to have lock-ups / crashes, even if > only occasionally. And crashing during a long video encoding job or game > session is definitely undesirable, never mind in a presentation. The > test often gives a quick result, and it's a definite result. > > Maybe Dell support would be more familiar with memtest & believe its result?
FWIW, I have found that memtest is a much faster tool for finding weak RAM than is Prime95. OTOH, Prime95 will find a weak CPU or motherboard that will never be revealed by memtest or any other torture test I've found (though I'm still looking). This goes for any type of system I've had to test from early Pentiums to my PE4400s. I'm also a fan of Prime95 for its ability to keep my processor at a constant temperature regardless of load. I believe this is one reason why my systems last longer than normal. Note that you will likely have to throttle Prime95 to get the best temperature regulation, else an increased system load will result in a *reduced* temperature (as all other processes are less stressful). The fact that it does something useful with my spare CPU cycles is just gravy. Bruce A Metcalf PrimeNet client #S00012 _______________________________________________ Prime mailing list [email protected] http://hogranch.com/mailman/listinfo/prime
