I don't have any links handy, though I'm sure a bit of googling would yield something.
The machine I built for my mother runs stable @ 2.3GHz. Unless a machine can run Prime95 for a while I don't think it's stable. F@H isn't a good way to see if a machine is stable because F@H doesn't die if it gets bad results, Prime95 does. Don't do F@H until you've made sure your machine passes 24h of Prime95 because F@H doesn't benefit from poor results, it just increases your score.
As for the CPU . . . I went ahead with the T-Bred(B) 0310XPMW DLT3C as it was $65 from ExcaliberPC, and I'm pretty hopeful (going by the results of the machine I built my mother) it will reach 2.3GHz @ 1.65v. The Barton was more expensive and I just didn't think it could reach 2.3GHz reliably.
I agree with you on the Athlon64. Buying a 130nm A64 probably isn't the best thing to do, especially w/ rumors of the 90nm A64 possibly using a difference socket (perhaps even the same as the Opteron).
Omega65 said There's a thread about a BARTON 3.0ghz (PR4200+) using a sub-zero Phase Change system, but Ihaven't heard of a Tbred hitting above 2.7 on straight water. If you have a link post it.
I also have yet to hear of a Tbreb or Barton above 2.4ghz running F@H 24/7 either. An CPUID pic of a CPU means little if you can't use it.
And don't feel bad about my OC efforts, Other than Thrax's Tbred 2.3 or Seversphere's Tbred 2.4, ( I know they run F@H) I take all reports about Tbred OC's with a CUP of salt.
My Barton 2.33 under water remained stable when the ambient room temp went from 70F to 88F. I have serious doubts that highend air would have kept a system folding the way mine did.
Hear's the advice..
1) Get a Barton 2500+
2) Forget the Athlon 64 until the 90nm versions arrive in 2004
If your CPU isn't fully stable, F@H will drop SSE optimizations on a machine that run P95 just fine. It will also fail to complete a WU if your CPU isn't fully stable.
That it doesn't crash just means the F@H programmers did a better job dealing with CPU failures than the P95 people
Join the S&M F@H effort and these issues will become apparent to you in time.
a2jfreak said Unless a machine can run Prime95 for a while I don't think it's stable. F@H isn't a good way to see if a machine is stable because F@H doesn't die if it gets bad results, Prime95 does. Don't do F@H until you've made sure your machine passes 24h of Prime95 because F@H doesn't benefit from poor results, it just increases your score.
Thats not true unfortunatly. Folding @ Home and especially the Gromacs is MUCH heavier on the cpu than Prime 95. Doesn´t matter really. As long as you can run your apps and games without problems, it´s stable. Prime 95 isn´t the sexiest proggy in the world either. Theres a lot of factors coming in to get a proggy to run stable, not only hardware. I know a lot of peeps that gets error in Prime at low speed when they can run it for days at a much higher speed. We use hardware that is a few weeks old, Prime 95 is VERY old.
Well, I've never had F@H die before, but I have had Prime95 die.
Prime95 tests against known good value, F@H does not.
So, please, tell me how F@H is better, since it doesn't stop when an error occurs (since it doesn't know) than Prime95 which does stop because it does know?
(I'm not talking about running Prime95 to search for new primes, I'm talking about the Toture Test.)
Please correct me where I'm wrong.
Mackanz said Thats not true unfortunatly. Folding @ Home and especially the Gromacs is MUCH heavier on the cpu than Prime 95. Doesn´t matter really. As long as you can run your apps and games without problems, it´s stable. Prime 95 isn´t the sexiest proggy in the world either. Theres a lot of factors coming in to get a proggy to run stable, not only hardware. I know a lot of peeps that gets error in Prime at low speed when they can run it for days at a much higher speed. We use hardware that is a few weeks old, Prime 95 is VERY old.
Omega65 said If your CPU isn't fully stable, F@H will drop SSE optimizations on a machine that run P95 just fine. It will also fail to complete a WU if your CPU isn't fully stable.
That it doesn't crash just means the F@H programmers did a better job dealing with CPU failures than the P95 people
Join the S&M F@H effort and these issues will become apparent to you in time.
F@H has never crashed on me either, but I've had plenty of Work Units fail to complete (because of errors). Gromac WUs are very sensiitive to overclocked CPUs.
TheWeatherMan said What is StressCPU?
This is a small windows program to torture-test your CPU in order to make sure that you don't have overheating problems. It will only run on SSE-equipped x86 CPUs, and it is executing a special version of the Gromacs innerloops that mixes SSE and normal assembly instructions to heat your CPU as much as possible.
The program was written by Erik Lindahl and I simply compiled it and am making it available to use.
This program actually makes my CPU's run from 4C-6C hotter than simply running Gromacs.
It's a good heat test and it should make any system draw maximum power which will also test the stability of your powersupply.
Let me know if you have any issues.
You can get the program in the EM-DC download area.
danball1976 said I'm not sure if I should wait until the Athlon64 comes out and the nForce3 as well, or wait for another nForce2 revision and the Barton
If I had that attitude I'd be trying to post this message from my dads old Commodore 64!
Prof
LOL, or even an old system using a Intel 4004 and 32KB of RAM.
heck...posting from a Tandy TRS-80 as i type....lol
i can say ive never tried Epox boards, Asus seems uninspiring, Abit....well Abit is Abit, i have two NF7-S's rev 2.0, a couple KX7-333 a couple KR7A's hanging around, some KT7's, a KT7A....all never gave a moments problem. as for MSI, its another favorite, it might not be a wicked overclocker, but i have to say that the KT3 ultra2 series are some really stable boards and i used two of'em until i got my first NF7-S.
i even got a friend of mine to trade his Asus A7V8X in for an Abit NF7-S, and hes lovin every minute of it, hes says the two dont even compare.
danball1976 said I'm not sure if I should wait until the Athlon64 comes out and the nForce3 as well, or wait for another nForce2 revision and the Barton
If I had that attitude I'd be trying to post this message from my dads old Commodore 64!
Prof
LOL, or even an old system using a Intel 4004 and 32KB of RAM.
heck...posting from a Tandy TRS-80 as i type....lol
Wow, how does that work? How much RAM does it have and what processor?
polarys425 said i can say ive never tried Epox boards, Asus seems uninspiring, Abit....well Abit is Abit, i have two NF7-S's rev 2.0, a couple KX7-333 a couple KR7A's hanging around, some KT7's, a KT7A....all never gave a moments problem. as for MSI, its another favorite, it might not be a wicked overclocker, but i have to say that the KT3 ultra2 series are some really stable boards and i used two of'em until i got my first NF7-S.
i even got a friend of mine to trade his Asus A7V8X in for an Abit NF7-S, and hes lovin every minute of it, hes says the two dont even compare.
I've owned both Epox and Abit brands. Abit BP-6, VH6-II, VH6-T, KT7, KT7A-R (3), KD7-R, KD7-E. For me Epox gives you the same feature set at a lower price. Abit tends to have better performance and a higher price. And Due to optimazations, Abit tends Not to be upward compatible with new Athlon Cores. (KT7A & Palominos, KR7 & Tbreds, KX7 & Tbreds(early models))
Comments
Thrax: 1
Shorty: 1
Omega65: 1 ( killer of motherboards
All tied up!
The machine I built for my mother runs stable @ 2.3GHz. Unless a machine can run Prime95 for a while I don't think it's stable. F@H isn't a good way to see if a machine is stable because F@H doesn't die if it gets bad results, Prime95 does. Don't do F@H until you've made sure your machine passes 24h of Prime95 because F@H doesn't benefit from poor results, it just increases your score.
As for the CPU . . . I went ahead with the T-Bred(B) 0310XPMW DLT3C as it was $65 from ExcaliberPC, and I'm pretty hopeful (going by the results of the machine I built my mother) it will reach 2.3GHz @ 1.65v. The Barton was more expensive and I just didn't think it could reach 2.3GHz reliably.
I agree with you on the Athlon64. Buying a 130nm A64 probably isn't the best thing to do, especially w/ rumors of the 90nm A64 possibly using a difference socket (perhaps even the same as the Opteron).
That it doesn't crash just means the F@H programmers did a better job dealing with CPU failures than the P95 people
Join the S&M F@H effort and these issues will become apparent to you in time.
Thrax & Mmonnin can chime in on this one.
Thats not true unfortunatly. Folding @ Home and especially the Gromacs is MUCH heavier on the cpu than Prime 95. Doesn´t matter really. As long as you can run your apps and games without problems, it´s stable. Prime 95 isn´t the sexiest proggy in the world either. Theres a lot of factors coming in to get a proggy to run stable, not only hardware. I know a lot of peeps that gets error in Prime at low speed when they can run it for days at a much higher speed. We use hardware that is a few weeks old, Prime 95 is VERY old.
Let me know when you're going to upgrade again, so I can buy stock in MB futures!:D
Prof
(The Smart-Alec):p
Prime95 tests against known good value, F@H does not.
So, please, tell me how F@H is better, since it doesn't stop when an error occurs (since it doesn't know) than Prime95 which does stop because it does know?
(I'm not talking about running Prime95 to search for new primes, I'm talking about the Toture Test.)
Please correct me where I'm wrong.
try this program
heck...posting from a Tandy TRS-80 as i type....lol
i even got a friend of mine to trade his Asus A7V8X in for an Abit NF7-S, and hes lovin every minute of it, hes says the two dont even compare.
Wow, how does that work? How much RAM does it have and what processor?
Still overall I like them both.