View Full Version : seti as a stability test
01-21-07, 01:37 PM
will it work? just wanna know so orthos doesnt cut into my SETI time when it doesnt have to.
01-21-07, 03:14 PM
If you are testing overclocks, then you may want to use something like Prime 95's torture test for stability issues instead of SETI or another distributed computing project. The torture test of P95 will compare your computer's results to know good results, which would help to tell you that your overclock is ok. SETI or other projects will not do that, and you may be sending those projects bad results that would set back their research, if they were used as stability testers. Hope that helps some. :)
01-21-07, 03:36 PM
well, as i said in the main post, i use orthos, but then the cpu time dedicated to seti drops to 0, and i dont like that :p
01-21-07, 06:16 PM
Well.. I guess you could use SETI as a stability test, though as Tyrinon said, it would be kinda detrimental towards the project's overall goal, if you were to send back "bad" results. Having said that, since the validators check your result with 2 others, I guess yours would get filtered pretty easily..
I guess then it comes down to preference. I personally wouldnt use SETI as a stability test, but I'm sure other people would disagree :P. It is after all, a nice CPU intensive app :)
01-21-07, 10:32 PM
Sorry about that. I overread that... :) I haven't used orthos, so how is it supposed to work?
01-22-07, 06:31 AM
basically, its just like prime95 only for multicore systems.
01-29-07, 11:13 AM
I still use Orthos to test basic stability - I just don't run it 24 hours anymore. I figure 4-9 hours depending on my schedule and let SETI test the long-range stuff. I still use a full 10 runs of MemTest for RAM though ...
vBulletin® v3.8.7, Copyright ©2000-2013, vBulletin Solutions, Inc.