Trials & Tribulations of Win7 SP1 Upgrade ...

[RonnyB]

I'm fairly new to F@H and am running 3 diff. Win7 rigs. This post regards my AMD Phenom X4 940 system:
- Gigabyte MA790X-UD4P mobo
- 4GB RAM
- EVGA GTS450SC
- EVGA GTS450
- 2 console clients v6.31
- HFM.net v0.4.8 rev121
- Process Lasso Pro v4.00.33
-Win7 Home Premium

I eventually got the SP1 upgrade to "stick" (after 4 or 5 tries at least) ... but it took help from some friends and a few phone calls with Microsoft support. Now the only problem is that the folding performance from the 2 GPUs is erratic ... sometimes at 99% + 99% ... and sometimes at 51% and 54% or sometimes 58% + 62%.

It the process of getting SP1 to actually complete its install, MS Support had me do an 'in place upgrade" (running Win7 install disk while in Win7 (still running)).

I had to reinstall the nVidia driver (v8.17.12.6658) after the "in place upgrade" and again after SP1 actually installed.

On my 3 machines I installed SP1 from the whole file (downloaded version from MS's site) and it was only on the AMD X4 940 machine that I had the problem. The failure would happen at the end when the machine rebooted and Sp1 was configuring ... after 2% or so it would say it had "failed" and was reverting.

I wound up (suggestion from a friend, not actually from MS Support) disabling almost everything in msconfig's StartUp and Processes tabs (except Virtual Disk). And another friend suggested temporarily disabling in Device Manager, one of my two Fermi cards ... "just in case". SP1 then stuck !

Anyway ever since, for several days, I have been getting this erratic usage of my 2 Fermi cards. Right now they are at 51% and 52%. A half hour ago they were at 99% and 99%.

Thanks for any insights !

[bruce]

I consider (99%+99%) to be "normal" for a Dual which isn't doing anything else. The other 1% or 2% can easily go to other minor activities including the GPU clients.

When it's getting (51% + 54%) or (58% + 62%), what is being reported as taking up the balance of your processors? Is it going to the "System Idle process" or to something else? (Be sure you have clicked "Show processes from all users")

[RonnyB]

Prior to doing the "in place upgrade" and the SP1 install, my AMD 940 was running 24/7 for weeks ... and GPU usage was always 99% + 99%.

Just to clarify: There is no FAH CPU/SMP client involved in this machine.

I know that Task Manager will show CPU usage ... but I am using EVGA Precision to show both GPU's usages.

[bruce]

Sorry. You were talking about GPU utilization....I interpreted it as CPU utilization. My mistake.

Are you still running the same GPU driver version?

[RonnyB]

Yes, am now running nVidia driver v8.17.12.6658 ... as I was before. At one point after the "inplace upgrade" and SP1 fiasco, I found that I had somehow reverted to the previous version of nVidia driver. I then soon upgraded to the current version and deleted the older driver install files from my downloads directory.

[RonnyB]

OK, I have bitten the bullet and tonight have blown away my 2 GPU clients (and their entire directories), taking the opportunity to download/install the newer v641 versions.
So far I see little improvement ... right now the 2 new console clients are running at 58% + 62% ... with PPD being a little better than 1/2 what it used to be in the good old days

So if it wasn't corrupted (somehow mysteriously by the "inplace upgrade" and SP1 install) GPU clients ... then what could it be?

The only thing I can think of now is to uninstall the nVidia driver, run Driver Cleaner Pro, and download the latest driver from nVidia ... again.

Or is there something else that could be done ?

[RonnyB]

I have the feeling that something must have borked my folding ability "slightly" when I did the "in place upgrade" and/or SP1 install.
Might it be worthwhile to uninstall all Folding related clients and support apps, and nuke any reference to them in the registry ... and then reinstall the folding apps, clients?

I will post my GPU console client's log info later this evening.

[PantherX]

I would suggest that you use Driver Sweeper to remove the drivers and install them again. Maybe a stubborn/corrupted file is the culprit.

[RonnyB]

Yes, using Driver Cleaner Pro on any nVidia fragments would be worth a try.

In the meantime, here is the FAHlog.txt from my 2 GTS450's;

Here's GPU1:

Code: Select all

--- Opening Log file [March 17 20:19:57 UTC] 


# Windows GPU Console Edition #################################################
###############################################################################

                       Folding@Home Client Version 6.41r2

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Folding\GPU-1
Executable: C:\Folding\GPU-1\FAH-GPU641.exe
Arguments: -verbosity 9 -local -advmethods -forceasm -gpu 0 

[20:19:57] - Ask before connecting: No
[20:19:57] - User name: Ron_S (Team 37412)
[20:19:57] - User ID: 69A192C23ECE447F
[20:19:57] - Machine ID: 1
[20:19:57] 
[20:19:57] Gpu type=3 species=21.
[20:19:57] Loaded queue successfully.
[20:19:57] 
[20:19:57] + Processing work unit
[20:19:57] Core required: FahCore_15.exe
[20:19:57] - Autosending finished units... [March 17 20:19:57 UTC]
[20:19:57] Core found.
[20:19:57] Trying to send all finished work units
[20:19:57] + No unsent completed units remaining.
[20:19:57] - Autosend completed
[20:19:57] Working on queue slot 03 [March 17 20:19:57 UTC]
[20:19:57] + Working ...
[20:19:57] - Calling '.\FahCore_15.exe -dir work/ -suffix 03 -nice 19 -nocpulock -checkpoint 15 -forceasm -verbose -lifeline 4164 -version 641'

[20:19:57] 
[20:19:57] *------------------------------*
[20:19:57] Folding@Home GPU Core
[20:19:57] Version 2.15 (Tue Nov 16 09:05:18 PST 2010)
[20:19:57] 
[20:19:57] Build host: SimbiosNvdWin7
[20:19:57] Board Type: NVIDIA/CUDA
[20:19:57] Core      : x=15
[20:19:57]  Window's signal control handler registered.
[20:19:57] Preparing to commence simulation
[20:19:57] - Assembly optimizations manually forced on.
[20:19:57] - Not checking prior termination.
[20:19:57] sizeof(CORE_PACKET_HDR) = 512 file=<>
[20:19:57] - Expanded 43672 -> 169787 (decompressed 388.7 percent)
[20:19:57] Called DecompressByteArray: compressed_data_size=43672 data_size=169787, decompressed_data_size=169787 diff=0
[20:19:57] - Digital signature verified
[20:19:57] 
[20:19:57] Project: 6800 (Run 5533, Clone 0, Gen 37)
[20:19:57] 
[20:19:57] Assembly optimizations on if available.
[20:19:57] Entering M.D.
[20:19:59] Will resume from checkpoint file work/wudata_03.ckp
[20:19:59] Tpr hash work/wudata_03.tpr:  271597542 1936794739 1852279528 2718501908 3792287603
[20:19:59] Working on PEPTIDE (1-42)
[20:19:59] Client config found, loading data.
[20:20:00] Starting GUI Server
[20:20:00] Resuming from checkpoint
[20:20:00] fcCheckPointResume: retreived and current tpr file hash:
[20:20:00]    0    271597542    271597542
[20:20:00]    1   1936794739   1936794739
[20:20:00]    2   1852279528   1852279528
[20:20:00]    3   2718501908   2718501908
[20:20:00]    4   3792287603   3792287603
[20:20:00] fcCheckPointResume: file hashes same.
[20:20:00] fcCheckPointResume: state restored.
[20:20:00] fcCheckPointResume: name work/wudata_03.log Verified work/wudata_03.log
[20:20:00] fcCheckPointResume: name work/wudata_03.trr Verified work/wudata_03.trr
[20:20:00] fcCheckPointResume: name work/wudata_03.xtc Verified work/wudata_03.xtc
[20:20:00] fcCheckPointResume: name work/wudata_03.edr Verified work/wudata_03.edr
[20:20:00] fcCheckPointResume: state restored 2
[20:20:00] Resumed from checkpoint
[20:20:00] Setting checkpoint frequency: 500000
[20:20:00] Completed   7500001 out of 50000000 steps (15%).
[20:23:47] Completed   8000000 out of 50000000 steps (16%).
[20:28:03] Completed   8500000 out of 50000000 steps (17%).
[20:31:50] Completed   9000000 out of 50000000 steps (18%).

and here's GPU2

Code: Select all

--- Opening Log file [March 17 20:36:20 UTC] 


# Windows GPU Console Edition #################################################
###############################################################################

                       Folding@Home Client Version 6.41r2

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Folding\GPU-2
Executable: C:\Folding\GPU-2\FAH-GPU641.exe
Arguments: -verbosity 9 -local -advmethods -forceasm -gpu 1 

[20:36:20] - Ask before connecting: No
[20:36:20] - User name: Ron_S (Team 37412)
[20:36:20] - User ID: 69A192C23ECE447F
[20:36:20] - Machine ID: 14
[20:36:20] 
[20:36:20] Gpu type=3 species=21.
[20:36:21] Loaded queue successfully.
[20:36:21] 
[20:36:21] + Processing work unit
[20:36:21] - Autosending finished units... [March 17 20:36:21 UTC]
[20:36:21] Core required: FahCore_15.exe
[20:36:21] Trying to send all finished work units
[20:36:21] Core found.
[20:36:21] + No unsent completed units remaining.
[20:36:21] - Autosend completed
[20:36:21] Working on queue slot 03 [March 17 20:36:21 UTC]
[20:36:21] + Working ...
[20:36:21] - Calling '.\FahCore_15.exe -dir work/ -suffix 03 -nice 19 -nocpulock -checkpoint 15 -forceasm -verbose -lifeline 3040 -version 641'

[20:36:21] 
[20:36:21] *------------------------------*
[20:36:21] Folding@Home GPU Core
[20:36:21] Version 2.15 (Tue Nov 16 09:05:18 PST 2010)
[20:36:21] 
[20:36:21] Build host: SimbiosNvdWin7
[20:36:21] Board Type: NVIDIA/CUDA
[20:36:21] Core      : x=15
[20:36:21]  Window's signal control handler registered.
[20:36:21] Preparing to commence simulation
[20:36:21] - Ensuring status. Please wait.
[20:36:31] - Assembly optimizations manually forced on.
[20:36:31] - Not checking prior termination.
[20:36:31] sizeof(CORE_PACKET_HDR) = 512 file=<>
[20:36:31] - Expanded 43443 -> 169787 (decompressed 390.8 percent)
[20:36:31] Called DecompressByteArray: compressed_data_size=43443 data_size=169787, decompressed_data_size=169787 diff=0
[20:36:31] - Digital signature verified
[20:36:31] 
[20:36:31] Project: 6800 (Run 11601, Clone 0, Gen 36)
[20:36:31] 
[20:36:31] Assembly optimizations on if available.
[20:36:31] Entering M.D.
[20:36:33] Will resume from checkpoint file work/wudata_03.ckp
[20:36:33] Tpr hash work/wudata_03.tpr:  2412111896 3449184296 3544130082 2447824741 4163099835
[20:36:33] Working on PEPTIDE (1-42)
[20:36:33] Client config found, loading data.
[20:36:33] Starting GUI Server
[20:36:33] Resuming from checkpoint
[20:36:33] fcCheckPointResume: retreived and current tpr file hash:
[20:36:33]    0   2412111896   2412111896
[20:36:33]    1   3449184296   3449184296
[20:36:33]    2   3544130082   3544130082
[20:36:33]    3   2447824741   2447824741
[20:36:33]    4   4163099835   4163099835
[20:36:33] fcCheckPointResume: file hashes same.
[20:36:33] fcCheckPointResume: state restored.
[20:36:33] fcCheckPointResume: name work/wudata_03.log Verified work/wudata_03.log
[20:36:33] fcCheckPointResume: name work/wudata_03.trr Verified work/wudata_03.trr
[20:36:33] fcCheckPointResume: name work/wudata_03.xtc Verified work/wudata_03.xtc
[20:36:34] fcCheckPointResume: name work/wudata_03.edr Verified work/wudata_03.edr
[20:36:34] fcCheckPointResume: state restored 2
[20:36:34] Resumed from checkpoint
[20:36:34] Setting checkpoint frequency: 500000
[20:36:34] Completed   7000001 out of 50000000 steps (14%).
[20:40:16] Completed   7500000 out of 50000000 steps (15%).
[20:43:58] Completed   8000000 out of 50000000 steps (16%).
[20:47:39] Completed   8500000 out of 50000000 steps (17%).
[20:51:21] Completed   9000000 out of 50000000 steps (18%).
[20:55:03] Completed   9500000 out of 50000000 steps (19%).
[20:58:44] Completed  10000000 out of 50000000 steps (20%).
[21:02:26] Completed  10500000 out of 50000000 steps (21%).
[21:06:08] Completed  11000000 out of 50000000 steps (22%).
[21:09:49] Completed  11500000 out of 50000000 steps (23%).
[21:13:30] Completed  12000000 out of 50000000 steps (24%).
[21:17:12] Completed  12500000 out of 50000000 steps (25%).
[21:20:54] Completed  13000000 out of 50000000 steps (26%).
[21:24:36] Completed  13500000 out of 50000000 steps (27%).
[21:28:17] Completed  14000000 out of 50000000 steps (28%).
[21:31:59] Completed  14500000 out of 50000000 steps (29%).
[21:35:42] Completed  15000000 out of 50000000 steps (30%).
[21:39:23] Completed  15500000 out of 50000000 steps (31%).
[21:43:05] Completed  16000000 out of 50000000 steps (32%).
[21:46:48] Completed  16500000 out of 50000000 steps (33%).
[21:50:29] Completed  17000000 out of 50000000 steps (34%).
[21:54:12] Completed  17500000 out of 50000000 steps (35%).
[21:57:54] Completed  18000000 out of 50000000 steps (36%).
[22:01:36] Completed  18500000 out of 50000000 steps (37%).
[22:05:19] Completed  19000000 out of 50000000 steps (38%).
[22:09:01] Completed  19500000 out of 50000000 steps (39%).
[22:12:42] Completed  20000000 out of 50000000 steps (40%).
[22:16:25] Completed  20500000 out of 50000000 steps (41%).
[22:20:06] Completed  21000000 out of 50000000 steps (42%).
[22:23:48] Completed  21500000 out of 50000000 steps (43%).
[22:27:31] Completed  22000000 out of 50000000 steps (44%).
[22:31:12] Completed  22500000 out of 50000000 steps (45%).
[22:34:55] Completed  23000000 out of 50000000 steps (46%).
[22:38:37] Completed  23500000 out of 50000000 steps (47%).
[22:42:19] Completed  24000000 out of 50000000 steps (48%).
[22:46:03] Completed  24500000 out of 50000000 steps (49%).
[22:49:46] Completed  25000000 out of 50000000 steps (50%).
[22:53:28] Completed  25500000 out of 50000000 steps (51%).
[22:57:11] Completed  26000000 out of 50000000 steps (52%).
[23:00:52] Completed  26500000 out of 50000000 steps (53%).
[23:04:35] Completed  27000000 out of 50000000 steps (54%).
[23:08:16] Completed  27500000 out of 50000000 steps (55%).
[23:11:57] Completed  28000000 out of 50000000 steps (56%).
[23:15:41] Completed  28500000 out of 50000000 steps (57%).
[23:19:23] Completed  29000000 out of 50000000 steps (58%).
[23:23:04] Completed  29500000 out of 50000000 steps (59%).
[23:26:47] Completed  30000000 out of 50000000 steps (60%).
[23:30:28] Completed  30500000 out of 50000000 steps (61%).
[23:34:11] Completed  31000000 out of 50000000 steps (62%).
[23:37:54] Completed  31500000 out of 50000000 steps (63%).
[23:41:35] Completed  32000000 out of 50000000 steps (64%).
[23:45:18] Completed  32499999 out of 50000000 steps (65%).
[23:49:00] Completed  32999999 out of 50000000 steps (66%).
[23:52:41] Completed  33499999 out of 50000000 steps (67%).
[23:56:23] Completed  33999999 out of 50000000 steps (68%).
[00:00:05] Completed  34499999 out of 50000000 steps (69%).
[00:03:47] Completed  34999999 out of 50000000 steps (70%).
[00:07:29] Completed  35499999 out of 50000000 steps (71%).
[00:11:11] Completed  35999999 out of 50000000 steps (72%).
[00:14:53] Completed  36499999 out of 50000000 steps (73%).
[00:18:35] Completed  36999999 out of 50000000 steps (74%).
[00:22:17] Completed  37499999 out of 50000000 steps (75%).
[00:26:00] Completed  37999999 out of 50000000 steps (76%).
[00:29:41] Completed  38499999 out of 50000000 steps (77%).
[00:33:23] Completed  38999999 out of 50000000 steps (78%).
[00:37:06] Completed  39499999 out of 50000000 steps (79%).
[00:40:47] Completed  39999999 out of 50000000 steps (80%).
[00:44:30] Completed  40499999 out of 50000000 steps (81%).
[00:48:12] Completed  40999999 out of 50000000 steps (82%).
[00:51:54] Completed  41499999 out of 50000000 steps (83%).
[00:55:36] Completed  41999999 out of 50000000 steps (84%).
[00:59:17] Completed  42499999 out of 50000000 steps (85%).
[01:02:59] Completed  42999999 out of 50000000 steps (86%).
[01:06:42] Completed  43499999 out of 50000000 steps (87%).
[01:10:23] Completed  43999999 out of 50000000 steps (88%).
[01:14:05] Completed  44499999 out of 50000000 steps (89%).
[01:17:47] Completed  44999999 out of 50000000 steps (90%).
[01:21:28] Completed  45499999 out of 50000000 steps (91%).
[01:25:12] Completed  45999999 out of 50000000 steps (92%).
[01:28:54] Completed  46499999 out of 50000000 steps (93%).
[01:32:35] Completed  46999999 out of 50000000 steps (94%).
[01:36:18] Completed  47499999 out of 50000000 steps (95%).
[01:40:00] Completed  47999999 out of 50000000 steps (96%).
[01:43:44] Completed  48499999 out of 50000000 steps (97%).
[01:47:27] Completed  48999999 out of 50000000 steps (98%).
[01:51:10] Completed  49499999 out of 50000000 steps (99%).
[01:54:54] Completed  49999999 out of 50000000 steps (100%).
[01:54:54] Finished fah_main
[01:54:54] 
[01:54:54] Successful run
[01:54:54] DynamicWrapper: Finished Work Unit: sleep=10000
[01:55:04] Reserved 2439964 bytes for xtc file; Cosm status=0
[01:55:04] Allocated 2439964 bytes for xtc file
[01:55:04] - Reading up to 2439964 from "work/wudata_03.xtc": Read 2439964
[01:55:04] Read 2439964 bytes from xtc file; available packet space=783990500
[01:55:04] xtc file hash check passed.
[01:55:04] Reserved 75840 75840 783990500 bytes for arc file=<work/wudata_03.trr> Cosm status=0
[01:55:04] Allocated 75840 bytes for arc file
[01:55:04] - Reading up to 75840 from "work/wudata_03.trr": Read 75840
[01:55:04] Read 75840 bytes from arc file; available packet space=783914660
[01:55:04] trr file hash check passed.
[01:55:04] Allocated 544 bytes for edr file
[01:55:04] Read bedfile
[01:55:04] edr file hash check passed.
[01:55:04] Allocated 121260 bytes for logfile
[01:55:04] Read logfile
[01:55:04] GuardedRun: success in DynamicWrapper
[01:55:04] GuardedRun: done
[01:55:04] Run: GuardedRun completed.
[01:55:08] + Opened results file
[01:55:08] - Writing 2638120 bytes of core data to disk...
[01:55:09] Done: 2637608 -> 2478875 (compressed to 93.9 percent)
[01:55:09]   ... Done.
[01:55:09] DeleteFrameFiles: successfully deleted file=work/wudata_03.ckp
[01:55:10] Shutting down core 
[01:55:10] 
[01:55:10] Folding@home Core Shutdown: FINISHED_UNIT
[01:55:13] CoreStatus = 64 (100)
[01:55:13] Unit 3 finished with 97 percent of time to deadline remaining.
[01:55:13] Updated performance fraction: 0.974062
[01:55:13] Sending work to server
[01:55:13] Project: 6800 (Run 11601, Clone 0, Gen 36)
[01:55:13] - Read packet limit of 540015616... Set to 524286976.

Thanks for any insights (spotting something that might be causing my "52% + 58%" problem

[RonnyB]

Please see the 4 screen captures of EVGA Precision below. They show the typical "irregularity" of F@H performance/GPU % of utilization (of my two GTS450's) ... that I have been experiencing.

What do the wavy lines ... when even , 96% or so is being acheived, mean/indicate ? I do not see this waviness on my other machine with 2 GTX260's.

Thanks

[bruce]

Is this the same problem: viewtopic.php?p=179446#p179446

[HendricksSA]

And is this after the clean and reinstall of drivers (what version)?

[RonnyB]

Is this the same problem: viewtopic.php?p=179446#p179446

It does sound similar ... but I have another AMD box, performing fine, running a single Fermi (GTX460) with SP1 and same Nvidia driver version (26658).

[RonnyB]

And is this after the clean and reinstall of drivers (what version)?

Yes, this is after ... well, the "inplace upgrade" followed by SP1 ... and re-install of the same driver (26658).

The funny thing is that everything was working fine before the "inplace upgrade" and SP1 update.

[HendricksSA]

Reinstall AND clean?

Also, I missed the Process Lasso entry in your OP. Is this being used in conjunction with the GPUs? (edited and added)