• Notice regarding image hosts and attachments. See here.

Belated parp @ Senture

Very odd. It's not something random like QPI link speed set to slow mode or something like that is it?
I hadn't heard about -smp7 issues, I'll bear that in mind if odd things start happening!
 
I have no idea mate, I'm working now but when I finish I'll be sprucing up weebeastie with its RMA'ed and replaced goodies. Although I must say I am loath to stop it foldinging a 60K bigadv :eek:

But once I've done that, I'll be attacking the main machine to see what I can do. Curious thing is that all the other benchmark apps I have run give it a clean bill of health, just folding@home is super slow on it.

But I'll be:

Swapping the GTX470 with the 480 in weebeastie.
Updating the bios
Setting everything to stock
Testing
If successful upping the clocks till at least 3.6Ghz and stable/cool
If not successful it'll be uninstall and reinstall stanfords apps.
If that doesn't work I'll try booting it off a clean install of Win7 on another drive see if that helps.

Can't think of much else. It was doing ok for a while then the 2.2 app came out and I tried a bigadv on it and it was so rubbish at that I tried all sorts of tweaks to get the performance up. I remember when I tried Vmware bigadv folding on it that it was no good at that either, but it always seemed to handle regular SMP work units in the past.

Never mind I'll get there even if I have to sell the motherboard and start afresh.
 
If all the performance tests show the hardware stacking up as it is, then it really must be software related. Have a stab at doing a clean install before dismantling hardware, since you water cool the beast it would make life easier if you don't have to break the loop.

Oh and Mr miniyazz, unless you have a secret weapon hidden that is going to save you, my bigadv unit is due to complete in 4.5 hours. Consider yourself parped :p
 
I have to swap the gfx card anyway :) the 480 pulls too many volts through the mobo on weebeastie. And tbh a clean install would be 10x more work, all that reinstalling software :eek:
 
Oh yeah, so much work. Windows 7 - 20 mins, drivers - 10 mins, windows updates - 15 mins, GPU overclocking - 5 mins, folding@home clients - 10 mins.

Yep, that sounds like lots of work :p
 
Senture said:
If all the performance tests show the hardware stacking up as it is, then it really must be software related. Have a stab at doing a clean install before dismantling hardware, since you water cool the beast it would make life easier if you don't have to break the loop.

Oh and Mr miniyazz, unless you have a secret weapon hidden that is going to save you, my bigadv unit is due to complete in 4.5 hours. Consider yourself parped :p
Click to expand...

Uh-oh. I shall get digging my secret re-entry passageway ;)

On the plus side, I'm on 353 pointers for my GPUs and a P6704 on the CPU and my PPD is now reported as 27,839.8 PPD :D
 
How embarrassing :o

In other news, when I switched to -smp 4 to do a bit of gaming, it decided to delete my half-done WU and start again :(
 
That's what happened with me and bigadv at first

Code: 
[09:17:44] Project: 2684 (Run 14, Clone 11, Gen 3)
[09:17:44] 
[09:17:44] Assembly optimizations on if available.
[09:17:44] Entering M.D.
[09:17:58] Completed 0 out of 250000 steps  (0%)
[10:06:14] Completed 2500 out of 250000 steps  (1%)
[10:54:23] Completed 5000 out of 250000 steps  (2%)
[11:42:36] Completed 7500 out of 250000 steps  (3%)
[12:30:45] Completed 10000 out of 250000 steps  (4%)
[13:19:30] Completed 12500 out of 250000 steps  (5%)
[14:07:39] Completed 15000 out of 250000 steps  (6%)
[14:55:41] Completed 17500 out of 250000 steps  (7%)
[15:43:24] Completed 20000 out of 250000 steps  (8%)
[16:31:20] Completed 22500 out of 250000 steps  (9%)
[17:20:01] Completed 25000 out of 250000 steps  (10%)
[17:26:03] Service stop request received.

Folding@Home Client Shutdown.

[17:30:35] Project: 2684 (Run 14, Clone 11, Gen 3)
[17:30:35] 
[17:30:37] Assembly optimizations on if available.
[17:30:37] Entering M.D.
[17:30:43] Using Gromacs checkpoints
[17:30:50] fcSaveRestoreState: I/O failed dir=0, var=0761F544, varsize=20
[17:30:50] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:50] fcSaveRestoreState: I/O failed dir=0, var=0775F544, varsize=20
[17:30:50] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:50] fcSaveRestoreState: I/O failed dir=0, var=0715F544, varsize=20
[17:30:50] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:51] fcSaveRestoreState: I/O failed dir=0, var=0705F544, varsize=20
[17:30:51] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:51] fcSaveRestoreState: I/O failed dir=0, var=0725F544, varsize=20
[17:30:51] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:53] fcSaveRestoreState: I/O failed dir=0, var=0739F544, varsize=20
[17:30:53] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:53] Resuming from checkpoint
[17:30:54] fcSaveRestoreState: I/O failed dir=0, var=074DF544, varsize=20
[17:30:54] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:30:54] fcSaveRestoreState: I/O failed dir=0, var=052DC5EC, varsize=20
[17:30:54] fcCheckPointResume: failure in call to fcSaveRestoreState() to restore cpt hash.
[17:31:29] CoreStatus = FF (255)
[17:31:29] Sending work to server
[17:31:29] Project: 2684 (Run 14, Clone 11, Gen 3)
[17:31:29] - Error: Could not get length of results file work/wuresults_02.dat
[17:31:29] - Error: Could not read unit 02 file. Removing from queue.
[17:31:29] - Preparing to get new work unit...
[17:31:29] Cleaning up work directory

Then got the same unit again and completed it. Now for some unknown reason, it's doing a normal A3 unit dropping PPD by 2k :(. Not to worry, at least it's not another useless P6702 :o
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom