is seti as good a torture test as prime 95?

smp

Diamond Member
Dec 6, 2000
5,215
0
76
is it? Cause seti runs the cpu 100%.. so is prime 95 necessary? I heard it rounds numbers or something if something is up.. let me know.. thanks.
 

Quad

Golden Member
Nov 18, 2000
1,222
0
0
well, i've personally never tried seti. i hear rc5 is pretty good too.
but if u want the ULTIMATE stress test, then i suggest you use the one concocted by Leo V. it may be extreme, but it's well worth it. you can check it out in Leo V's review of the a7v133 here.

here's a copy of the stress test:
quote:
The final problem which I've recently solved is stability. After all the pain with random lockups, I wanted nothing short of 100% reliability. Everything looked good at first, until I invented a new level of PC torture. All of these were run simultaneously:

* Huge directory copy over 100mbps LAN (2 simultaneous copy processes)
* CPU burn-in program with error checking
* CoolEdit upsampling to a huge WAV file, CPU and disk-intensive process
* Playing a CD-quality or 88KHz/32-bit (to parasite bandwidth) WAV file over the LAN
* Exact Audio Copy top-quality audio CD ripping/testing to harddisk
* Random IE windows/browsing
* Outlook Express open
* Random explorer windows



hehe good luck
 

borealiss

Senior member
Jun 23, 2000
913
0
0
just because a program uses all the cpu doesn't mean that it is a good stress test. the code segment "while(1)" will use the cpu all the way, but it is nowhere as good as seti or rc5. and those should be tested in conjunction with other programs as well. rc5 is mainly integer, and seti does fourier transforms which is fpu intensive, so each will stress a different part of the cpu. get these two puppies going, maybe defrag a slave drive with garbage on it, play a fps game in a window or run a quake 3 loop, and do normal browsing and stuff. sometimes i scan 100 meg images in photoshop, run a filter, defrag, run q3a in a window with max nightmare bots, run rc5, seti, and play 5 or so streaming divx movies and mp3's. the idea is to stress as many of your i/o subsystems as possible, ie graphics, disk, etc...
 

smp

Diamond Member
Dec 6, 2000
5,215
0
76
Yeah.. I read LeoV's original post there.. I liked it.. but yeah, I have my own version.. open 3dsmax, photoshop, director, flash, autocad all at the same time.. while, playing a mp3, compressing a bunch of mp3's and copying a movie from one partition to the other..
passed at an fsb of 103, failed at fsb 105.. but I think I failed that second time because my hhd was on FIRE!!! :|
but yeah, seti was on all the time too.. I think I would have passed had it not been for the hdd being on fire.. copying, compressing and opening a tonne of sh!t really made it hot.. and I don't have a fan on it or anything.. like almost burned me to the touch.. ouch!
But yeah, I havn't unlocked my cpu yet, cause I can't get the fvckin heatsink off.. I have the stock tbird heatsink on there right now.. so I have a 45 mhz overclock now.. what I really wanted to test.. was my ram.. cause I have it at cas 2 138mhz..
 

esung

Golden Member
Oct 13, 1999
1,063
0
0
SETI@home will not report errors induced by the CPU, or it couldn't detect the error. but Prime95 has a tendency to catch even the slightest error if you OC the CPU a little too much. People will argue that although Prime95 is report errors, the rest of the system still feels fine. If you do a search, you'll see quite a few I think...

Personally, I believe what Prime95 is telling me. :D