How to lower F@H use of GPU?

chengbin · Post by **chengbin** » Sun May 31, 2009 2:43 am

Is there a way to make F@H GPU to only use say like 50% of my GPU's capability?

I'm asking because after 1 year of torture from F@H GPU and CPU, maxing the CPU and GPU usage for 16 hours a day for a year, my power supply and motherboard died. It was on a Friday, which sucks because technicians don't work on weekends, and I must wait till Monday for them to come to my house to fix it. It is also right into my summative/exam period, making it even more frustrating.

I have a Q6600 + 8800GT (both not overclocked) running on a 425 watt PSU. Is that enough for constant stress that I've been doing for the last year?

MoneyGuyBK · Post by **MoneyGuyBK** » Sun May 31, 2009 2:58 am

I believe this was discussed in another thread.
If you use the SysTray client then you could use the slider to adjust usage.
However with regards to the Console client this is not yet available.

Peace

7im · Post by **7im** » Sun May 31, 2009 3:02 am

On a quality PSU, it's fine, and should run many years. On a consumer level box like an entry level Dell computer, it's a crap shoot. Might run a year, might run 10.

In the Systray Client, there is a usage slider, but not sure how well that works. I think it reduces the CPU usage, which then starves the GPU of data, thus lowering the GPU usage. Some have said it works to help reduce heat on some of the larger GPU work units.

Post by **bruce** » Sun May 31, 2009 4:07 am

As far as I know there has been no systematic testing of what happens when you reduce the %CPU setting.

For a particular WU, somebody could report what CPU settings of 10%, 20%, 30% ... 80%, 90%, 100% does to the %GPU values. They're probably not one-to-one. Once that table of values is reported, it should be repeated for other proteins. Whether the same table of values would apply to a protein with a different number of atoms is questionable so until it is tested, nobody knows.

Also, if you're running CPU clients or SMP clients concurrently, you may choose to violate the recommendation that the GPU client should be at a higher priority than the CPU/SMP client. That would also reduce the throughput of the GPU by some unknown amount. Be aware that there would still be brief periods of high GPU utilization when the CPU/SMP client is between WUs, uploading and downloading, so this is probably not as useful as the CPU slider.

Any reports would be appreciated if you're interested in testing it.

planetclown · Post by **planetclown** » Sun May 31, 2009 11:24 am

MoneyGuyBK wrote:I believe this was discussed in another thread.
If you use the SysTray client then you could use the slider to adjust usage.
However with regards to the Console client this is not yet available.

The console asks what percentage of the CPU to use. Doesn't this do the same thing as the SysTray slider, only static?

Post by **toTOW** » Sun May 31, 2009 2:27 pm

chengbin wrote:I have a Q6600 + 8800GT (both not overclocked) running on a 425 watt PSU. Is that enough for constant stress that I've been doing for the last year?

That's what happens with under-dimensioned PSUs ... a good quality 500W PSU is a minimum for your set up in my opinion.

Dark Light · Post by **Dark Light** » Mon Jun 01, 2009 1:34 am

Is there a way to restrict the GPU load on a 8800 GTS to about 95%, so that the OS doesn't seem so slow?

patonb · Post by **patonb** » Mon Jun 01, 2009 1:53 am

In the systray config, you can.

7im · Post by **7im** » Mon Jun 01, 2009 1:53 am

Already a topic running on this... Use the slider in the Systray client, and turn on Aero.

http://foldingforum.org/viewtopic.php?f=16&t=10180

Edit by Mod:
Threads merged
.b

7im · Post by **7im** » Mon Jun 01, 2009 2:07 am

And don't forget, the total Watts of a PSU is not the primary factor to consider. Yes, you need a minimum wattage to run a certain GPU, but also a minimum Amperage. Plus you don't want to run the PSU at 90% usage. You want some headroom to upgrade or add components later.

What I use as a guide, is to calculate what I expect the total Watts and Amps to be, and then double it (and always round up to the next highest wattage). Then that's the PSU I buy. PSU is not overly taxed, fans run quiet, PSU runs inside it's higher efficiency band, I have headroom to upgrade, and I don't blow too much cash on an over-spec'd PSU, nor do I burn them out in only a year. And if it doesn't come with a 5 year warranty (in the US), it ain't worth the gamble to me. Some data is too precious to risk on inexpensive hardware.

Dark Light · Post by **Dark Light** » Mon Jun 01, 2009 2:20 am

Thanks its seemed to have worked, I thought that was to restricted the CPU load.

MoneyGuyBK · Post by **MoneyGuyBK** » Mon Jun 01, 2009 4:06 am

planetclown wrote:
MoneyGuyBK wrote:I believe this was discussed in another thread.
If you use the SysTray client then you could use the slider to adjust usage.
However with regards to the Console client this is not yet available.
The console asks what percentage of the CPU to use. Doesn't this do the same thing as the SysTray slider, only static?

This method will not really work with the Console version ... by adjusting the CPU usage you are only limiting the usage of CPU during uploads of results and downloads of projects ... and THAT only happens every 30-300 Minutes depending on the size of the Work Unit and the performance of your Video Card.... the video card usage remains at 100%

If you must control the usage of your GPU, then switch to and use the SysTray version until PG finds a way to remedy this issue.
If you have no heat issues or display issues I would recommend you not worry about this.

Peace

Amaruk · Post by **Amaruk** » Wed Jun 03, 2009 8:40 am

MoneyGuyBK wrote:
planetclown wrote:
MoneyGuyBK wrote:I believe this was discussed in another thread.
If you use the SysTray client then you could use the slider to adjust usage.
However with regards to the Console client this is not yet available.
The console asks what percentage of the CPU to use. Doesn't this do the same thing as the SysTray slider, only static?

This method will not really work with the Console version ... by adjusting the CPU usage you are only limiting the usage of CPU during uploads of results and downloads of projects ... and THAT only happens every 30-300 Minutes depending on the size of the Work Unit and the performance of your Video Card.... the video card usage remains at 100%

If you must control the usage of your GPU, then switch to and use the SysTray version until PG finds a way to remedy this issue.
If you have no heat issues or display issues I would recommend you not worry about this.

Peace

I've tested both the console and systray GPU clients. Both of them work the same. The slider (systray) and % (console) both throttle the Fahcore, reducing the load on the GPU.

Bruce is correct in that the relation between setting and actual usage is not linear.

I'm not home atm, l'll post more details when I get back.

Amaruk · Post by **Amaruk** » Mon Jun 08, 2009 8:24 pm

This test was done with 6.23 Console client using 185.85 drivers

Project: 5753 (Run 0, Clone 43, Gen 241)

GPU - 8800 GT 256MB G92 running at 700/750/1700 (3rd card)

MB - MSI K9A2 Platinum

CPU - X2 5600 @ 2.9 Ghz

Mem - 4 gb PC2-6400

OS - XP Pro SP3 32bit

Every user process except Fahcore_11.exe was locked to core 0. The FahCore was locked to 1.

Client configured to run on third GPU. No monitor/dummy attached. Used -forcegpu flag.

Ran client to 10% complete to get baseline. 3344.73 PPD, 68C

Set CPU usage to 25% and ran from 10 until 20% complete. 1236.71 PPD at 56C

Set CPU usage to 50% and ran from 20 until 38% complete. 2072.79 PPD at 61C

Set CPU usage to 75% and ran from 38 until 56% complete. 2904.63 PPD at 63C

Set CPU usage to 100% and ran from 56 until 67% complete. 3344.73 PPD at 65C

Restarted at 67% completion with -oneunit to finish.

Here is a composite of Task Manager and Fahmon for each step in CPU usage.

The Fahcore is the only process on core 1 and the graph confirms that the CPU usage setting controls it instead of the client.

It's unclear if the CPU usage control affects the GPU directly or if the CPU itself controls the GPU.

Either way the result is the same. The reduced GPU usage is reflected by Fahmon.

Of interest is the relationship between CPU% and PPD.

At 25% the PPD is 1236.71, which is 36.97% of 3344.73

At 50% the PPD is 2072.79, which is 61.97% of 3344.73

At 75% the PPD is 2904.63, which is 86.84% of 3344.73

These numbers are almost an exact match with the Systray test I ran earlier.

Variation in PPD drop as related to CPU usage averaged about 2% between Console and Systray.

Here is the logfile:

Code: Select all


--- Opening Log file [June 3 05:17:47 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -configonly 

[05:17:47] Configuring Folding@Home...


[05:22:04] - Ask before connecting: No
[05:22:04] - User name: Amaruk (Team 50625)
[05:22:04] - User ID: 67CFDE0276C8CD35
[05:22:04] - Machine ID: 2
[05:22:04] 
[05:22:04] -configonly flag given, so exiting.


--- Opening Log file [June 3 05:22:21 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[05:22:21] - Ask before connecting: No
[05:22:21] - User name: Amaruk (Team 50625)
[05:22:21] - User ID: 67CFDE0276C8CD35
[05:22:21] - Machine ID: 2
[05:22:21] 
[05:22:21] Work directory not found. Creating...
[05:22:21] Could not open work queue, generating new queue...
[05:22:21] - Preparing to get new work unit...
[05:22:21] - Autosending finished units... [June 3 05:22:21 UTC]
[05:22:21] + Attempting to get work packet
[05:22:21] Trying to send all finished work units
[05:22:21] + No unsent completed units remaining.
[05:22:21] - Will indicate memory of 2815 MB
[05:22:21] - Autosend completed
[05:22:21] - Detect CPU. Vendor: AuthenticAMD, Family: 15, Model: 11, Stepping: 2
[05:22:21] - Connecting to assignment server
[05:22:21] Connecting to http://assign-GPU.stanford.edu:8080/
[05:22:28] Posted data.
[05:22:28] Initial: 43AB; - Successful: assigned to (171.67.108.11).
[05:22:28] + News From Folding@Home: Welcome to Folding@Home
[05:22:28] Loaded queue successfully.
[05:22:28] Connecting to http://171.67.108.11:8080/
[05:22:29] Posted data.
[05:22:29] Initial: 0000; - Receiving payload (expected size: 97030)
[05:22:29] Conversation time very short, giving reduced weight in bandwidth avg
[05:22:29] - Downloaded at ~189 kB/s
[05:22:29] - Averaged speed for that direction ~189 kB/s
[05:22:29] + Received work.
[05:22:29] + Closed connections
[05:22:29] 
[05:22:29] + Processing work unit
[05:22:29] Core required: FahCore_11.exe
[05:22:29] Core not found.
[05:22:29] - Core is not present or corrupted.
[05:22:29] - Attempting to download new core...
[05:22:29] + Downloading new core: FahCore_11.exe
[05:22:29] Downloading core (/~pande/Win32/x86/NVIDIA/G80/Core_11.fah from www.stanford.edu)
[05:22:29] Initial: AFDE; + 10240 bytes downloaded
[05:22:29] Initial: 8CE4; + 20480 bytes downloaded
[05:22:29] Initial: D8CE; + 30720 bytes downloaded
[05:22:29] Initial: 268C; + 40960 bytes downloaded
[05:22:29] Initial: F040; + 51200 bytes downloaded
[05:22:29] Initial: 8F3D; + 61440 bytes downloaded
[05:22:29] Initial: E22C; + 71680 bytes downloaded
[05:22:29] Initial: 232D; + 81920 bytes downloaded
[05:22:30] Initial: 7B96; + 92160 bytes downloaded
[05:22:30] Initial: 35ED; + 102400 bytes downloaded
[05:22:30] Initial: 12C7; + 112640 bytes downloaded
[05:22:30] Initial: CE46; + 122880 bytes downloaded
[05:22:30] Initial: 4CD5; + 133120 bytes downloaded
[05:22:30] Initial: 0CD7; + 143360 bytes downloaded
[05:22:30] Initial: 0D37; + 153600 bytes downloaded
[05:22:30] Initial: 2D4A; + 163840 bytes downloaded
[05:22:30] Initial: 602C; + 174080 bytes downloaded
[05:22:30] Initial: 4001; + 184320 bytes downloaded
[05:22:30] Initial: D84F; + 194560 bytes downloaded
[05:22:30] Initial: 2151; + 204800 bytes downloaded
[05:22:30] Initial: 2ABF; + 215040 bytes downloaded
[05:22:30] Initial: 5CA8; + 225280 bytes downloaded
[05:22:30] Initial: 40D4; + 235520 bytes downloaded
[05:22:30] Initial: 1804; + 245760 bytes downloaded
[05:22:30] Initial: 1478; + 256000 bytes downloaded
[05:22:30] Initial: D0B4; + 266240 bytes downloaded
[05:22:30] Initial: B858; + 276480 bytes downloaded
[05:22:30] Initial: 1830; + 286720 bytes downloaded
[05:22:30] Initial: BA4E; + 296960 bytes downloaded
[05:22:30] Initial: 3985; + 307200 bytes downloaded
[05:22:30] Initial: 232C; + 317440 bytes downloaded
[05:22:30] Initial: 63DF; + 327680 bytes downloaded
[05:22:30] Initial: 0AAF; + 337920 bytes downloaded
[05:22:30] Initial: 7F28; + 348160 bytes downloaded
[05:22:30] Initial: 9005; + 358400 bytes downloaded
[05:22:30] Initial: 8084; + 368640 bytes downloaded
[05:22:30] Initial: FBCE; + 378880 bytes downloaded
[05:22:30] Initial: F76F; + 389120 bytes downloaded
[05:22:30] Initial: C594; + 399360 bytes downloaded
[05:22:30] Initial: B0BC; + 409600 bytes downloaded
[05:22:30] Initial: DD0D; + 419840 bytes downloaded
[05:22:30] Initial: DCF5; + 430080 bytes downloaded
[05:22:30] Initial: 6D67; + 440320 bytes downloaded
[05:22:30] Initial: A987; + 450560 bytes downloaded
[05:22:30] Initial: E0CA; + 460800 bytes downloaded
[05:22:30] Initial: 6090; + 471040 bytes downloaded
[05:22:30] Initial: 83FD; + 481280 bytes downloaded
[05:22:30] Initial: D0F9; + 491520 bytes downloaded
[05:22:30] Initial: A4B9; + 501760 bytes downloaded
[05:22:30] Initial: 7046; + 512000 bytes downloaded
[05:22:30] Initial: 3F00; + 522240 bytes downloaded
[05:22:30] Initial: 07F1; + 532480 bytes downloaded
[05:22:30] Initial: 6E7D; + 542720 bytes downloaded
[05:22:30] Initial: 4416; + 552960 bytes downloaded
[05:22:30] Initial: 353A; + 563200 bytes downloaded
[05:22:30] Initial: 6C4D; + 573440 bytes downloaded
[05:22:30] Initial: 305A; + 583680 bytes downloaded
[05:22:30] Initial: 3D23; + 593920 bytes downloaded
[05:22:30] Initial: A242; + 604160 bytes downloaded
[05:22:30] Initial: 1E02; + 614400 bytes downloaded
[05:22:30] Initial: 75E0; + 624640 bytes downloaded
[05:22:30] Initial: 8F3A; + 634880 bytes downloaded
[05:22:30] Initial: 1E1B; + 642475 bytes downloaded
[05:22:30] Verifying core Core_11.fah...
[05:22:30] Signature is VALID
[05:22:30] 
[05:22:30] Trying to unzip core FahCore_11.exe
[05:22:30] Decompressed FahCore_11.exe (1843200 bytes) successfully
[05:22:35] + Core successfully engaged
[05:22:40] 
[05:22:40] + Processing work unit
[05:22:40] Core required: FahCore_11.exe
[05:22:40] Core found.
[05:22:40] Working on queue slot 01 [June 3 05:22:40 UTC]
[05:22:40] + Working ...
[05:22:40] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -checkpoint 3 -verbose -lifeline 2400 -version 623'

[05:22:41] 
[05:22:41] *------------------------------*
[05:22:41] Folding@Home GPU Core - Beta
[05:22:41] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[05:22:41] 
[05:22:41] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[05:22:41] Build host: amoeba
[05:22:41] Board Type: Nvidia
[05:22:41] Core      : 
[05:22:41] Preparing to commence simulation
[05:22:41] - Looking at optimizations...
[05:22:41] - Created dyn
[05:22:41] - Files status OK
[05:22:41] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[05:22:41] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[05:22:41] - Digital signature verified
[05:22:41] 
[05:22:41] Project: 5753 (Run 0, Clone 43, Gen 241)
[05:22:41] 
[05:22:41] Assembly optimizations on if available.
[05:22:41] Entering M.D.
[05:22:47] Working on Protein
[05:22:50] Client config found, loading data.
[05:22:50] Starting GUI Server
[05:25:02] Completed 1%
[05:27:15] Completed 2%
[05:29:27] Completed 3%
[05:31:40] Completed 4%
[05:33:52] Completed 5%
[05:36:05] Completed 6%
[05:38:17] Completed 7%
[05:40:30] Completed 8%
[05:42:42] Completed 9%
[05:44:55] Completed 10%
[05:45:14] ***** Got a SIGTERM signal (2)
[05:45:14] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 05:45:20 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -configonly -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[05:45:20] - Ask before connecting: No
[05:45:20] - User name: Amaruk (Team 50625)
[05:45:20] - User ID: 67CFDE0276C8CD35
[05:45:20] - Machine ID: 2
[05:45:20] 
[05:45:20] Configuring Folding@Home...


[05:45:44] - Ask before connecting: No
[05:45:44] - User name: Amaruk (Team 50625)
[05:45:44] - User ID: 67CFDE0276C8CD35
[05:45:44] - Machine ID: 2
[05:45:44] 
[05:45:44] -configonly flag given, so exiting.


--- Opening Log file [June 3 05:45:51 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[05:45:51] - Ask before connecting: No
[05:45:51] - User name: Amaruk (Team 50625)
[05:45:51] - User ID: 67CFDE0276C8CD35
[05:45:51] - Machine ID: 2
[05:45:51] 
[05:45:51] Loaded queue successfully.
[05:45:51] 
[05:45:51] - Autosending finished units... [June 3 05:45:51 UTC]
[05:45:51] + Processing work unit
[05:45:51] Trying to send all finished work units
[05:45:51] Core required: FahCore_11.exe
[05:45:51] + No unsent completed units remaining.
[05:45:51] Core found.
[05:45:51] - Autosend completed
[05:45:51] Working on queue slot 01 [June 3 05:45:51 UTC]
[05:45:51] + Working ...
[05:45:51] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -cpu 25 -checkpoint 3 -verbose -lifeline 2644 -version 623'

[05:45:51] 
[05:45:51] *------------------------------*
[05:45:51] Folding@Home GPU Core - Beta
[05:45:51] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[05:45:51] 
[05:45:51] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[05:45:51] Build host: amoeba
[05:45:51] Board Type: Nvidia
[05:45:51] Core      : 
[05:45:51] Preparing to commence simulation
[05:45:51] - Looking at optimizations...
[05:45:51] - Files status OK
[05:45:51] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[05:45:51] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[05:45:51] - Digital signature verified
[05:45:51] 
[05:45:51] Project: 5753 (Run 0, Clone 43, Gen 241)
[05:45:51] 
[05:45:52] Assembly optimizations on if available.
[05:45:52] Entering M.D.
[05:45:58] Will resume from checkpoint file
[05:45:59] Working on Protein
[05:46:08] Client config found, loading data.
[05:46:08] Starting GUI Server
[05:46:09] Resuming from checkpoint
[05:46:09] Verified work/wudata_01.log
[05:46:09] Verified work/wudata_01.edr
[05:46:09] Verified work/wudata_01.xtc
[05:46:09] Completed 10%
[05:52:06] Completed 11%
[05:58:04] Completed 12%
[06:04:01] Completed 13%
[06:09:58] Completed 14%
[06:15:56] Completed 15%
[06:21:53] Completed 16%
[06:27:51] Completed 17%
[06:33:48] Completed 18%
[06:39:45] Completed 19%
[06:45:43] Completed 20%
[06:50:54] ***** Got a SIGTERM signal (2)
[06:50:54] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 06:51:00 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -configonly -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[06:51:00] - Ask before connecting: No
[06:51:00] - User name: Amaruk (Team 50625)
[06:51:00] - User ID: 67CFDE0276C8CD35
[06:51:00] - Machine ID: 2
[06:51:00] 
[06:51:00] Configuring Folding@Home...


[06:51:24] - Ask before connecting: No
[06:51:24] - User name: Amaruk (Team 50625)
[06:51:24] - User ID: 67CFDE0276C8CD35
[06:51:24] - Machine ID: 2
[06:51:24] 
[06:51:24] -configonly flag given, so exiting.


--- Opening Log file [June 3 06:51:29 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[06:51:29] - Ask before connecting: No
[06:51:29] - User name: Amaruk (Team 50625)
[06:51:29] - User ID: 67CFDE0276C8CD35
[06:51:29] - Machine ID: 2
[06:51:29] 
[06:51:29] Loaded queue successfully.
[06:51:29] 
[06:51:29] + Processing work unit
[06:51:29] - Autosending finished units... [06:51:29]
[06:51:29] Core required: FahCore_11.exe
[06:51:29] Trying to send all finished work units
[06:51:29] Core found.
[06:51:29] + No unsent completed units remaining.
[06:51:29] - Autosend completed
[06:51:29] Working on queue slot 01 [June 3 06:51:29 UTC]
[06:51:29] + Working ...
[06:51:29] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -cpu 50 -checkpoint 3 -verbose -lifeline 3388 -version 623'

[06:51:29] 
[06:51:29] *------------------------------*
[06:51:29] Folding@Home GPU Core - Beta
[06:51:29] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[06:51:29] 
[06:51:29] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[06:51:29] Build host: amoeba
[06:51:29] Board Type: Nvidia
[06:51:29] Core      : 
[06:51:29] Preparing to commence simulation
[06:51:29] - Looking at optimizations...
[06:51:29] - Files status OK
[06:51:29] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[06:51:29] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[06:51:29] - Digital signature verified
[06:51:29] 
[06:51:29] Project: 5753 (Run 0, Clone 43, Gen 241)
[06:51:29] 
[06:51:30] Assembly optimizations on if available.
[06:51:30] Entering M.D.
[06:51:36] Will resume from checkpoint file
[06:51:36] Working on Protein
[06:51:41] Client config found, loading data.
[06:51:41] Starting GUI Server
[06:51:41] Resuming from checkpoint
[06:51:41] Verified work/wudata_01.log
[06:51:41] Verified work/wudata_01.edr
[06:51:41] Verified work/wudata_01.xtc
[06:51:41] Completed 20%
[06:55:15] Completed 21%
[06:58:48] Completed 22%
[07:02:21] Completed 23%
[07:05:55] Completed 24%
[07:09:29] Completed 25%
[07:13:02] Completed 26%
[07:16:36] Completed 27%
[07:20:09] Completed 28%
[07:23:43] Completed 29%
[07:27:16] Completed 30%
[07:30:50] Completed 31%
[07:34:23] Completed 32%
[07:37:56] Completed 33%
[07:41:30] Completed 34%
[07:45:03] Completed 35%
[07:48:37] Completed 36%
[07:52:10] Completed 37%
[07:55:44] Completed 38%
[07:56:58] ***** Got a SIGTERM signal (2)
[07:56:58] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 07:57:03 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -configonly -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[07:57:03] - Ask before connecting: No
[07:57:03] - User name: Amaruk (Team 50625)
[07:57:03] - User ID: 67CFDE0276C8CD35
[07:57:03] - Machine ID: 2
[07:57:03] 
[07:57:03] Configuring Folding@Home...


[07:57:18] - Ask before connecting: No
[07:57:18] - User name: Amaruk (Team 50625)
[07:57:18] - User ID: 67CFDE0276C8CD35
[07:57:18] - Machine ID: 2
[07:57:18] 
[07:57:18] -configonly flag given, so exiting.


--- Opening Log file [June 3 07:57:24 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[07:57:24] - Ask before connecting: No
[07:57:24] - User name: Amaruk (Team 50625)
[07:57:24] - User ID: 67CFDE0276C8CD35
[07:57:24] - Machine ID: 2
[07:57:24] 
[07:57:24] Loaded queue successfully.
[07:57:24] 
[07:57:24] + Processing work unit
[07:57:24] - Autosending finished units... [June 3 07:57:24 UTC]
[07:57:24] Core required: FahCore_11.exe
[07:57:24] Trying to send all finished work units
[07:57:24] Core found.
[07:57:24] + No unsent completed units remaining.
[07:57:24] - Autosend completed
[07:57:24] Working on queue slot 01 [June 3 07:57:24 UTC]
[07:57:24] + Working ...
[07:57:24] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -cpu 75 -checkpoint 3 -verbose -lifeline 1576 -version 623'

[07:57:24] 
[07:57:24] *------------------------------*
[07:57:24] Folding@Home GPU Core - Beta
[07:57:24] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[07:57:24] 
[07:57:24] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[07:57:24] Build host: amoeba
[07:57:24] Board Type: Nvidia
[07:57:24] Core      : 
[07:57:24] Preparing to commence simulation
[07:57:24] - Looking at optimizations...
[07:57:24] - Files status OK
[07:57:24] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[07:57:24] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[07:57:24] - Digital signature verified
[07:57:24] 
[07:57:24] Project: 5753 (Run 0, Clone 43, Gen 241)
[07:57:24] 
[07:57:24] Assembly optimizations on if available.
[07:57:24] Entering M.D.
[07:57:30] Will resume from checkpoint file
[07:57:31] Working on Protein
[07:57:34] Client config found, loading data.
[07:57:34] Starting GUI Server
[07:57:34] Resuming from checkpoint
[07:57:34] Verified work/wudata_01.log
[07:57:34] Verified work/wudata_01.edr
[07:57:34] Verified work/wudata_01.xtc
[07:57:34] Completed 38%
[08:00:06] Completed 39%
[08:02:38] Completed 40%
[08:05:10] Completed 41%
[08:07:42] Completed 42%
[08:10:14] Completed 43%
[08:12:46] Completed 44%
[08:15:18] Completed 45%
[08:17:51] Completed 46%
[08:20:23] Completed 47%
[08:22:55] Completed 48%
[08:25:27] Completed 49%
[08:27:59] Completed 50%
[08:30:31] Completed 51%
[08:33:03] Completed 52%
[08:35:36] Completed 53%
[08:38:08] Completed 54%
[08:40:40] Completed 55%
[08:43:12] Completed 56%
[08:45:43] ***** Got a SIGTERM signal (2)
[08:45:43] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 08:45:48 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -configonly -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[08:45:48] - Ask before connecting: No
[08:45:48] - User name: Amaruk (Team 50625)
[08:45:48] - User ID: 67CFDE0276C8CD35
[08:45:48] - Machine ID: 2
[08:45:48] 
[08:45:48] Configuring Folding@Home...


[08:46:04] - Ask before connecting: No
[08:46:04] - User name: Amaruk (Team 50625)
[08:46:04] - User ID: 67CFDE0276C8CD35
[08:46:04] - Machine ID: 2
[08:46:04] 
[08:46:04] -configonly flag given, so exiting.


--- Opening Log file [June 3 08:46:09 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[08:46:09] - Ask before connecting: No
[08:46:09] - User name: Amaruk (Team 50625)
[08:46:09] - User ID: 67CFDE0276C8CD35
[08:46:09] - Machine ID: 2
[08:46:09] 
[08:46:09] Loaded queue successfully.
[08:46:09] 
[08:46:09] - Autosending finished units... [June 3 08:46:09 UTC]
[08:46:09] + Processing work unit
[08:46:09] Trying to send all finished work units
[08:46:09] Core required: FahCore_11.exe
[08:46:09] + No unsent completed units remaining.
[08:46:09] Core found.
[08:46:09] - Autosend completed
[08:46:09] Working on queue slot 01 [June 3 08:46:09 UTC]
[08:46:09] + Working ...
[08:46:09] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -checkpoint 3 -verbose -lifeline 3480 -version 623'

[08:46:09] 
[08:46:09] *------------------------------*
[08:46:09] Folding@Home GPU Core - Beta
[08:46:09] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[08:46:09] 
[08:46:09] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[08:46:09] Build host: amoeba
[08:46:09] Board Type: Nvidia
[08:46:09] Core      : 
[08:46:09] Preparing to commence simulation
[08:46:09] - Looking at optimizations...
[08:46:09] - Files status OK
[08:46:09] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[08:46:09] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[08:46:09] - Digital signature verified
[08:46:09] 
[08:46:09] Project: 5753 (Run 0, Clone 43, Gen 241)
[08:46:09] 
[08:46:09] Assembly optimizations on if available.
[08:46:09] Entering M.D.
[08:46:15] Will resume from checkpoint file
[08:46:15] Working on Protein
[08:46:18] Client config found, loading data.
[08:46:18] Starting GUI Server
[08:46:18] Resuming from checkpoint
[08:46:18] Verified work/wudata_01.log
[08:46:18] Verified work/wudata_01.edr
[08:46:18] Verified work/wudata_01.xtc
[08:46:18] Completed 56%
[08:48:30] Completed 57%
[08:50:43] Completed 58%
[08:52:55] Completed 59%
[08:55:08] Completed 60%
[08:57:20] Completed 61%
[08:59:33] Completed 62%
[09:01:45] Completed 63%
[09:03:58] Completed 64%
[09:06:10] Completed 65%
[09:08:23] Completed 66%
[09:10:35] Completed 67%
[09:11:09] ***** Got a SIGTERM signal (2)
[09:11:09] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 09:11:31 UTC] 


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

                       Folding@Home Client Version 6.23

                          http://folding.stanford.edu

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

Launch directory: D:\Console_GPU
Executable: D:\Console_GPU\[email protected]
Arguments: -oneunit -gpu 2 -verbosity 9 -forcegpu nvidia_g80 

[09:11:31] - Ask before connecting: No
[09:11:31] - User name: Amaruk (Team 50625)
[09:11:31] - User ID: 67CFDE0276C8CD35
[09:11:31] - Machine ID: 2
[09:11:31] 
[09:11:31] Loaded queue successfully.
[09:11:31] 
[09:11:31] - Autosending finished units... [June 3 09:11:31 UTC]
[09:11:31] + Processing work unit
[09:11:31] Trying to send all finished work units
[09:11:31] Core required: FahCore_11.exe
[09:11:31] + No unsent completed units remaining.
[09:11:31] Core found.
[09:11:31] - Autosend completed
[09:11:31] Working on queue slot 01 [June 3 09:11:31 UTC]
[09:11:31] + Working ...
[09:11:31] - Calling '.\FahCore_11.exe -dir work/ -suffix 01 -checkpoint 3 -verbose -lifeline 3152 -version 623'

[09:11:31] 
[09:11:31] *------------------------------*
[09:11:31] Folding@Home GPU Core - Beta
[09:11:31] Version 1.19 (Mon Nov 3 09:34:13 PST 2008)
[09:11:31] 
[09:11:31] Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[09:11:31] Build host: amoeba
[09:11:31] Board Type: Nvidia
[09:11:31] Core      : 
[09:11:31] Preparing to commence simulation
[09:11:31] - Looking at optimizations...
[09:11:31] - Files status OK
[09:11:31] - Expanded 96518 -> 489240 (decompressed 506.8 percent)
[09:11:31] Called DecompressByteArray: compressed_data_size=96518 data_size=489240, decompressed_data_size=489240 diff=0
[09:11:31] - Digital signature verified
[09:11:31] 
[09:11:31] Project: 5753 (Run 0, Clone 43, Gen 241)
[09:11:31] 
[09:11:31] Assembly optimizations on if available.
[09:11:31] Entering M.D.
[09:11:37] Will resume from checkpoint file
[09:11:38] Working on Protein
[09:11:40] Client config found, loading data.
[09:11:40] Starting GUI Server
[09:11:40] Resuming from checkpoint
[09:11:40] Verified work/wudata_01.log
[09:11:40] Verified work/wudata_01.edr
[09:11:40] Verified work/wudata_01.xtc
[09:11:40] Completed 67%
[09:13:53] Completed 68%
[09:16:05] Completed 69%
[09:18:17] Completed 70%
[09:20:30] Completed 71%
[09:22:42] Completed 72%
[09:24:54] Completed 73%
[09:27:07] Completed 74%
[09:29:19] Completed 75%
[09:31:31] Completed 76%
[09:33:44] Completed 77%
[09:35:56] Completed 78%
[09:38:08] Completed 79%
[09:40:21] Completed 80%
[09:42:33] Completed 81%
[09:44:45] Completed 82%
[09:46:58] Completed 83%
[09:49:10] Completed 84%
[09:51:22] Completed 85%
[09:53:35] Completed 86%
[09:55:47] Completed 87%
[09:57:59] Completed 88%
[10:00:12] Completed 89%
[10:02:24] Completed 90%
[10:04:36] Completed 91%
[10:06:49] Completed 92%
[10:09:01] Completed 93%
[10:11:13] Completed 94%
[10:13:26] Completed 95%
[10:15:38] Completed 96%
[10:17:50] Completed 97%
[10:20:03] Completed 98%
[10:22:15] Completed 99%
[10:24:27] Completed 100%
[10:24:28] Successful run
[10:24:28] DynamicWrapper: Finished Work Unit: sleep=10000
[10:24:38] Reserved 112228 bytes for xtc file; Cosm status=0
[10:24:38] Allocated 112228 bytes for xtc file
[10:24:38] - Reading up to 112228 from "work/wudata_01.xtc": Read 112228
[10:24:38] Read 112228 bytes from xtc file; available packet space=786318236
[10:24:38] xtc file hash check passed.
[10:24:38] Reserved 33528 33528 786318236 bytes for arc file=<work/wudata_01.trr> Cosm status=0
[10:24:38] Allocated 33528 bytes for arc file
[10:24:38] - Reading up to 33528 from "work/wudata_01.trr": Read 33528
[10:24:38] Read 33528 bytes from arc file; available packet space=786284708
[10:24:38] trr file hash check passed.
[10:24:38] Allocated 560 bytes for edr file
[10:24:38] Read bedfile
[10:24:38] edr file hash check passed.
[10:24:38] Allocated 32284 bytes for logfile
[10:24:38] Read logfile
[10:24:38] GuardedRun: success in DynamicWrapper
[10:24:38] GuardedRun: done
[10:24:38] Run: GuardedRun completed.
[10:24:41] - Writing 179112 bytes of core data to disk...
[10:24:41] Done: 178600 -> 155307 (compressed to 86.9 percent)
[10:24:41]   ... Done.
[10:24:42] - Shutting down core 
[10:24:42] 
[10:24:42] Folding@home Core Shutdown: FINISHED_UNIT
[10:24:45] CoreStatus = 64 (100)
[10:24:45] Unit 1 finished with 93 percent of time to deadline remaining.
[10:24:45] Updated performance fraction: 0.930031
[10:24:45] Sending work to server
[10:24:45] Project: 5753 (Run 0, Clone 43, Gen 241)


[10:24:45] + Attempting to send results [June 3 10:24:45 UTC]
[10:24:45] - Reading file work/wuresults_01.dat from core
[10:24:45]   (Read 155819 bytes from disk)
[10:24:45] Connecting to http://171.67.108.11:8080/
[10:24:46] Posted data.
[10:24:46] Initial: 0000; - Uploaded at ~153 kB/s
[10:24:46] - Averaged speed for that direction ~153 kB/s
[10:24:46] + Results successfully sent
[10:24:46] Thank you for your contribution to Folding@Home.
[10:24:46] + Starting local stats count at 1
[10:24:50] Trying to send all finished work units
[10:24:50] + No unsent completed units remaining.
[10:24:50] + -oneunit flag given and have now finished a unit. Exiting.***** Got a SIGTERM signal (2)
[10:24:50] Killing all core threads

Folding@Home Client Shutdown.

Post by **bruce** » Mon Jun 08, 2009 8:29 pm

Very interesting.

Thank you for your report and your systematic testing.

If you are assigned a new WU with a different number of atoms (and points) you might want to repeat selected parts of the tests. It's not certain that the numbers will be proportional.

Folding Forum

How to lower F@H use of GPU?

How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

How Do I Restrict GPU Load?

Re: How Do I Restrict GPU Load?

Re: How Do I Restrict GPU Load?

Re: How to lower F@H use of GPU?

Re: How Do I Restrict GPU Load?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?

Re: How to lower F@H use of GPU?