Soldato
Yep, never had a dualcore with 2 equal cores, hate when one core can go wayyyyy further than the other 
Prime95 Vs SuperPI
- Thread starter wdEvA
- Start date
I run prime95 for one iteration (about 10 minutes) and then folding perminantly after that. Not good practice really, i should run it for longer but i can't be bothered and i see no point really when it could be doing something usefull like folding.
If it can fold, its got to be stable enough to play a few games and whatnot.
Memory errors are a different story, i was running a machine for a while that would fail memtest in a big way every time and it passed prime and folded quite happily for a month or so, but it did crash in games.
So prime/sp2004 for a few hours, memtest and folding is my recomendation.
If it can fold, its got to be stable enough to play a few games and whatnot.
Memory errors are a different story, i was running a machine for a while that would fail memtest in a big way every time and it passed prime and folded quite happily for a month or so, but it did crash in games.
So prime/sp2004 for a few hours, memtest and folding is my recomendation.
I laugh at the people who claim that they won't accept that a PC is stable unless it has run Prime for at least 8/12/24 hours.
I have had rigs run Prime for as much as 16 hours and fall over after running F@H for a couple of hours - hell, I even had one run Prime for 8 hours and fall down in 1M superpi.
Prime is not the be all and end all of testing - what's important is that it is stable enough to perform the tasks that you require it to perform.
One of my Folding rigs failed Prime after 6 minutes, it has been running F@H at 100% on both cores for 6 months, pretty much 24/7, without any problems. That's all I need it to do. I couldn't care less if others don't consider it to be stable - it does what I want it to do and that's good enough for me.
The insistance on PCs being "Prime stable" is overclocking snobbery and very narrowminded.
Stan
I have had rigs run Prime for as much as 16 hours and fall over after running F@H for a couple of hours - hell, I even had one run Prime for 8 hours and fall down in 1M superpi.
Prime is not the be all and end all of testing - what's important is that it is stable enough to perform the tasks that you require it to perform.
One of my Folding rigs failed Prime after 6 minutes, it has been running F@H at 100% on both cores for 6 months, pretty much 24/7, without any problems. That's all I need it to do. I couldn't care less if others don't consider it to be stable - it does what I want it to do and that's good enough for me.
The insistance on PCs being "Prime stable" is overclocking snobbery and very narrowminded.
Stan
Soldato
Prime is just one test i use, its good mainly for the cpu, but memtest86 and 3dmark are also good stressers, keep laughing while i use my rock solid pcs.
I ran prime for 10 minutes and memtest and mine is just as stable as yours. It does not crash due to hardware problems and it folds constantly. So was it worth the hours and hours spent running a million iterations of prime?Justintime said:
I agree with stan on this one, either a machine is stable or it isn't, yours isn't 'more' stable than mine and therefore better than mine because you tested it for ages and i didn't.
Memtest is different, because memory errors can be extremely random and therefore in certain rare cases its worth running that for 24 hours, but to me running prime for more than a couple of hours is a waste of time unless your testing things other than stability, such as cooling ability.
Soldato
To each his own When i test with prime i am testing cpu stability, psu stability, cooling and to some extent RAM. It works for me and i feel its worth it and in the end thats all that matters.
When i test with prime i am testing cpu stability, psu stability, cooling and to some extent RAM. It works for me and i feel its worth it and in the end thats all that matters.
Associate
How do you set up Prime to work?
I am new to this overclocking stuff, and think I can push things further, but need to know how to test it properly!
Cheers
I am new to this overclocking stuff, and think I can push things further, but need to know how to test it properly!
Cheers
Justintime said:To each his own
Exactly my point.
You are happy that Prime95 is an adequate test of your machine's stability - who am I to say you are wrong, it works for you and that's what's important.
I put less store in Prime95 and prefer to judge stability in a system's ability to perform the tasks required of it (in my case, mostly Folding but I do also have a couple of gaming rigs) and that works for me - everybody's happy (except the overclocking snobs who won't accept that my rigs are stable unless I conform to their ideals).
Stan
There is no definitive test, although like many others I trust Prime95 the most.
The fact is a successfully overclocked CPU should be able to pass ANY test thrown at it (the same as it would at stock). Many people won't accept this for the simple reason it often means the difference between a 20% or a 40% overclock.
The way I look at overclocking is ANY extra speed is great, but I'm not going to snub stability tests just so I can parade a 40% overclock in my signature.
The fact is a successfully overclocked CPU should be able to pass ANY test thrown at it (the same as it would at stock). Many people won't accept this for the simple reason it often means the difference between a 20% or a 40% overclock.
The way I look at overclocking is ANY extra speed is great, but I'm not going to snub stability tests just so I can parade a 40% overclock in my signature.
Associate
Can anyone explain how to set up Prime to test properly?
When I run it my CPU only shows 50% usage on both cores!
Cheers guys
When I run it my CPU only shows 50% usage on both cores!
Cheers guys
Soldato
mmj_uk said:
I disagree.
A successfully overclocked CPU should be able to do what the user requires it to do regardless of whether it can pass some artificially designed test which has little relevance to everyday use.
Everyone to their own though - if you want to subject your hardware to those sort of tests to satisfy your need for "stability" then you go ahead. I'll carry on doing it my way (with 8 rigs Folding 24/7 - all overclocked - I must be doing something right
).Stan
Soldato
Looks like I am on the fence for this one because although some of the tests are synthetic - I still want my CPU to pass
I don't overclock every 10 minutes though due to having just the one rig at a time usually - so I can afford to spend a bit of time running Prime95 and a bit of Memtest.
Some people seem to think you need to put a 100m exclusion zone around your PC when Prime testing! Silly really because you can still use it - just don't do anything too CPU intensive!
I wouldn't honestly be happy if my OC died after less than 24 hours in Prime95. That is just me and my path to perfection
. Not saying my OC is more stable, but I may have a better idea of how stable my OC is!
SiriusB
I don't overclock every 10 minutes though due to having just the one rig at a time usually - so I can afford to spend a bit of time running Prime95 and a bit of Memtest.
Some people seem to think you need to put a 100m exclusion zone around your PC when Prime testing! Silly really because you can still use it - just don't do anything too CPU intensive!
I wouldn't honestly be happy if my OC died after less than 24 hours in Prime95. That is just me and my path to perfection
. Not saying my OC is more stable, but I may have a better idea of how stable my OC is! SiriusB