F@H - Tuesday, July 26

[dawks]

Howdy ladies

I'll be adding several P4 1.7Ghz systems soon

 

[brainwrinkle]

Is this what you're looking for?

[13:40:19] Folding@Home Gromacs Core
[13:40:19] Version 1.81 (April 6, 2005)
[13:40:19]
[13:40:19] Preparing to commence simulation
[13:40:19] - Looking at optimizations...
[13:40:19] - Created dyn
[13:40:19] - Files status OK
[13:40:20] + Screensaver: engaging
[13:40:20] - Expanded 564554 -> 2749385 (decompressed 487.0 percent)
[13:40:20] - Starting from initial work packet
[13:40:20]
[13:40:20] Project: 219 (Run 3, Clone 16, Gen 52)
[13:40:20]
[13:40:20] Assembly optimizations on if available.
[13:40:20] Entering M.D.
[13:40:26] Protein: p219_c21Murea
[13:40:26]
[13:40:26] Writing local files
[13:40:29] Extra SSE boost OK.
[13:40:29] Writing local files
[13:40:29] Completed 0 out of 1000000 steps (0)
[13:43:13] + Screensaver: no longer idle, ending
[13:59:52] Writing local files
[13:59:52] Completed 10000 out of 1000000 steps (1)
[14:19:35] Writing local files
[14:19:35] Completed 20000 out of 1000000 steps (2)
[14:39:22] Writing local files
[14:39:22] Completed 30000 out of 1000000 steps (3)
[14:59:25] Writing local files
[14:59:25] Completed 40000 out of 1000000 steps (4)

Looks to me like every frame has 1000 steps.

I got the normal points and frames from F@H Current Projects

 

[brainwrinkle]

I also posted my problem at F@H forums, and they figured it out. Each project has the same number of steps, but can have a different number of frames. Each one of my frames had 1000 steps, while the benchmark computer had 10000 step frames. That is how I had 1000 frames instead of 100.

 

[rise]

i'm sure some of you started folding as a stability test- how would you know if your system is unstable by using this? a heat issue? crashes? or the results take too long?

 

[kb3edk]

Originally posted by: rise4310
i'm sure some of you started folding as a stability test- how would you know if your system is unstable by using this? a heat issue? crashes? or the results take too long?

When my system was unstable, the simulation (not the client) would crash about 3 or 4 frames into a Gromacs WU, then it would try to submit the results, and download a new WU, sim would then crash again after a couple frames, etc. There's a limit that you can only crash the sim 5 times, then the client gets put on pause so you can "fix" the problem.

 

[rise]

cool, thanks. i'm still learning how this all works so when i come across the 2-3 day WU's i just want to make sure i'm not supplying bad results or doing anything wrong. i mean, i'm prime stable for more than 24 hours but thats just a simple code.

 

[Insidious]

Sounds like you're doing great!

Keep up the good work

-Sid

PS: Unstability can be a lot of different things... BSOD, work units failing (you would see it mentioned in your FahLog.txt file), etc.... Basically, if you crunch full time and don't have anything unusual happen, you can consider yourself stable... same as with Prime95.

If I were to notice crashes or 'wierdities' I would make Step 1 of my troubleshooting returning volts and clocks to stock. Folding (or Prime, etc.) doesn't make you unstable, it just lets subtle issues be noticed more quickly since you are executing all of the time.

 

[Insidious]

Originally posted by: dawks
Howdy ladies

I'll be adding several P4 1.7Ghz systems soon

:beer: :thumbsup: :beer:

-Sid

 

[JeffCos]

Originally posted by: dawks
Howdy ladies

I'll be adding several P4 1.7Ghz systems soon

Great, as if enough people haven't been passing me lately...