Page 1 of 1
Sanity check: Is this proper Kepler throttling?
Posted: Sat Mar 02, 2013 12:02 am
by coredump4
I got a great deal on an open-box EVGA GTX 650 "SuperClocked" 1GB GDDR5 board recently, so I wanted to test out its Folding performance relative to the GTX 560 Ti I also have. Not to mention, Folding is a great burn-in test for any used or open-box gear.
My initial test went fine in terms of stability, but the 650's performance was about half of the 560 Ti. I was expecting the two to be more similar, largely due to having the same number of shaders. But since the 650 is my first exposure to Kepler, I just want a sanity check of what I'm seeing. When looking around for clues, I saw that the 650 might have been throttling during folding. I've heard the Keplers can do this, but I got mixed information from the EVGA "NV-Z" monitoring tool. Here's the 650 when idle; notice the Current and Max GPU core clocks in the P8 & P0 states. And as expected, the Current PState's GPU clock matches the "GPU clock" value below:
However, while folding, look what happens not only to the Max GPU clock, but that the Current PState's GPU clock doesn't match the GPU clock shown below:
So this leads me to think 1 of 3 things is happening:
1) The GPU is throttling, but if so, why the apparent discrepancies in the reported current core clock?
2) The NV-Z tool and/or GeForce driver are reporting incorrect clock data;
3) Both 1) & 2) are happening.
Environment: WinXP SP3 32-bit, GeForce 314.07 driver, FAHclient 7.3.6.
Thanks for any insight into this behavior.
Re: Sanity check: Is this proper Kepler throttling?
Posted: Sun Mar 03, 2013 4:32 am
by PantherX
Have you tried using a different monitoring tool like GPU-Z to verify the finding of EVGA? Make sure that you don't run more than one monitoring tool at a time as it may conflict with another and cause issues.
Also, please post the Project number and the Time Per Frame (TPF) so that others with the same GPU can look at it and comment if it is within the acceptable range or not. If you are not sure how to do it, please post the log here within the Code Tags (in the Full Editor).
Re: Sanity check: Is this proper Kepler throttling?
Posted: Sun Mar 03, 2013 10:47 am
by Nathan_P
Given that NV-Z is only picking up 1MB of GDDR 5 i would say that it is not reading properly, As panther suggests - download and run GPU-Z
Re: Sanity check: Is this proper Kepler throttling?
Posted: Mon Mar 04, 2013 9:43 pm
by coredump4
I loaded GPU-Z on that system, but haven't swapped the 650 back in yet, but I'll do that in the next day or two and report back. Though I will say, on another system I have, every tool reports the card's RAM improperly except GPU-Z...
It is project 7626. I don't know how to properly calculate TPF from the log, though if a "step" in the log corresponds to a fold then it would be easy. Anyway, here's the log... note that FAHclient refers to the GTX 650 as a "GT 650":
Code: Select all
*********************** Log Started 2013-02-26T14:31:29Z ***********************
14:31:29:************************* Folding@home Client *************************
14:31:29: Website: http://folding.stanford.edu/
14:31:29: Copyright: (c) 2009-2013 Stanford University
14:31:29: Author: Joseph Coffland <[email protected]>
14:31:29: Args:
14:31:29: Config: C:/Documents and Settings/redacted/Application
14:31:29: Data/FAHClient/config.xml
14:31:29:******************************** Build ********************************
14:31:29: Version: 7.3.6
14:31:29: Date: Feb 18 2013
14:31:29: Time: 15:25:17
14:31:29: SVN Rev: 3923
14:31:29: Branch: fah/trunk/client
14:31:29: Compiler: Intel(R) C++ MSVC 1500 mode 1200
14:31:29: Options: /TP /nologo /EHa /Qdiag-disable:4297,4103,1786,279 /Ox -arch:SSE
14:31:29: /QaxSSE2,SSE3,SSSE3,SSE4.1,SSE4.2 /Qopenmp /Qrestrict /MT /Qmkl
14:31:29: Platform: win32 XP
14:31:29: Bits: 32
14:31:29: Mode: Release
14:31:29:******************************* System ********************************
14:31:29: CPU: Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
14:31:29: CPU ID: GenuineIntel Family 6 Model 23 Stepping 10
14:31:29: CPUs: 2
14:31:29: Memory: 3.25GiB
14:31:29: Free Memory: 2.79GiB
14:31:29: Threads: WINDOWS_THREADS
14:31:29: Has Battery: false
14:31:29: On Battery: false
14:31:29: UTC offset: -5
14:31:29: PID: 1880
14:31:29: CWD: C:/Documents and Settings/redacted/Application Data/FAHClient
14:31:29: OS: Microsoft Windows XP Service Pack 3
14:31:29: OS Arch: X86
14:31:29: GPUs: 1
14:31:29: GPU 0: NVIDIA:3 GK107 [GeForce GT 650]
14:31:29: CUDA: 3.0
14:31:29: CUDA Driver: 5000
14:31:29:Win32 Service: false
14:31:29:***********************************************************************
14:31:29:<config>
14:31:29: <!-- Folding Slot Configuration -->
14:31:29: <power v='full'/>
14:31:29: <smp v='false'/>
14:31:29:
14:31:29: <!-- Network -->
14:31:29: <proxy v=':8080'/>
14:31:29:
14:31:29: <!-- User Information -->
14:31:29: <user v='BRZfan'/>
14:31:29:
14:31:29: <!-- Folding Slots -->
14:31:29: <slot id='1' type='GPU'/>
14:31:29:</config>
14:31:29:Trying to access database...
14:31:29:Successfully acquired database lock
14:31:29:Enabled folding slot 01: READY gpu:0:GK107 [GeForce GT 650]
14:31:29:WU00:FS01:Starting
14:31:29:WU00:FS01:Running FahCore: "C:\Program Files\FAHClient/FAHCoreWrapper.exe" "C:/Documents and Settings/redacted/Application Data/FAHClient/cores/www.stanford.edu/~pande/Win32/x86/NVIDIA/Fermi/Core_15.fah/FahCore_15.exe" -dir 00 -suffix 01 -version 703 -lifeline 1880 -checkpoint 15 -gpu 0 -gpu-vendor nvidia
14:31:29:WU00:FS01:Started FahCore on PID 1580
14:31:29:WU00:FS01:Core PID:1544
14:31:29:WU00:FS01:FahCore 0x15 started
14:31:29:WU00:FS01:0x15:
14:31:29:WU00:FS01:0x15:*------------------------------*
14:31:29:WU00:FS01:0x15:Folding@Home GPU Core
14:31:29:WU00:FS01:0x15:Version 2.25 (Wed May 9 17:03:01 EDT 2012)
14:31:29:WU00:FS01:0x15:Build host AmoebaRemote
14:31:29:WU00:FS01:0x15:Board Type NVIDIA/CUDA
14:31:29:WU00:FS01:0x15:Core 15
14:31:29:WU00:FS01:0x15:
14:31:29:WU00:FS01:0x15:Window's signal control handler registered.
14:31:29:WU00:FS01:0x15:Preparing to commence simulation
14:31:29:WU00:FS01:0x15:- Ensuring status. Please wait.
14:31:38:WU00:FS01:0x15:- Looking at optimizations...
14:31:38:WU00:FS01:0x15:- Working with standard loops on this execution.
14:31:38:WU00:FS01:0x15:- Previous termination of core was improper.
14:31:38:WU00:FS01:0x15:- Files status OK
14:31:38:WU00:FS01:0x15:sizeof(CORE_PACKET_HDR) = 512 file=<>
14:31:38:WU00:FS01:0x15:- Expanded 126190 -> 507182 (decompressed 401.9 percent)
14:31:38:WU00:FS01:0x15:Called DecompressByteArray: compressed_data_size=126190 data_size=507182, decompressed_data_size=507182 diff=0
14:31:38:WU00:FS01:0x15:- Digital signature verified
14:31:38:WU00:FS01:0x15:
14:31:38:WU00:FS01:0x15:Project: 7626 (Run 629, Clone 3, Gen 15)
14:31:38:WU00:FS01:0x15:
14:31:38:WU00:FS01:0x15:Entering M.D.
14:31:40:WU00:FS01:0x15:Will resume from checkpoint file 00/wudata_01.ckp
14:31:40:WU00:FS01:0x15:Tpr hash 00/wudata_01.tpr: 3088256186 2730024186 1947096957 1214887156 2277528353
14:31:40:WU00:FS01:0x15:GPU device id=0
14:31:40:WU00:FS01:0x15:Working on Protein
14:31:40:WU00:FS01:0x15:Client config unavailable.
14:31:41:WU00:FS01:0x15:Starting GUI Server
14:32:10:FS01:Paused
14:32:10:FS01:Shutting core down
14:32:18:WU00:FS01:FahCore returned: INTERRUPTED (102 = 0x66)
14:39:50:FS01:Unpaused
14:39:50:WU00:FS01:Starting
14:39:50:WU00:FS01:Running FahCore: "C:\Program Files\FAHClient/FAHCoreWrapper.exe" "C:/Documents and Settings/redacted/Application Data/FAHClient/cores/www.stanford.edu/~pande/Win32/x86/NVIDIA/Fermi/Core_15.fah/FahCore_15.exe" -dir 00 -suffix 01 -version 703 -lifeline 1880 -checkpoint 15 -gpu 0 -gpu-vendor nvidia
14:39:50:WU00:FS01:Started FahCore on PID 2040
14:39:50:WU00:FS01:Core PID:3060
14:39:50:WU00:FS01:FahCore 0x15 started
14:39:51:WU00:FS01:0x15:
14:39:51:WU00:FS01:0x15:*------------------------------*
14:39:51:WU00:FS01:0x15:Folding@Home GPU Core
14:39:51:WU00:FS01:0x15:Version 2.25 (Wed May 9 17:03:01 EDT 2012)
14:39:51:WU00:FS01:0x15:Build host AmoebaRemote
14:39:51:WU00:FS01:0x15:Board Type NVIDIA/CUDA
14:39:51:WU00:FS01:0x15:Core 15
14:39:51:WU00:FS01:0x15:
14:39:51:WU00:FS01:0x15:Window's signal control handler registered.
14:39:51:WU00:FS01:0x15:Preparing to commence simulation
14:39:51:WU00:FS01:0x15:- Looking at optimizations...
14:39:51:WU00:FS01:0x15:- Files status OK
14:39:51:WU00:FS01:0x15:sizeof(CORE_PACKET_HDR) = 512 file=<>
14:39:51:WU00:FS01:0x15:- Expanded 126190 -> 507182 (decompressed 401.9 percent)
14:39:51:WU00:FS01:0x15:Called DecompressByteArray: compressed_data_size=126190 data_size=507182, decompressed_data_size=507182 diff=0
14:39:51:WU00:FS01:0x15:- Digital signature verified
14:39:51:WU00:FS01:0x15:
14:39:51:WU00:FS01:0x15:Project: 7626 (Run 629, Clone 3, Gen 15)
14:39:51:WU00:FS01:0x15:
14:39:51:WU00:FS01:0x15:Assembly optimizations on if available.
14:39:51:WU00:FS01:0x15:Entering M.D.
14:39:52:WU00:FS01:0x15:Will resume from checkpoint file 00/wudata_01.ckp
14:39:52:WU00:FS01:0x15:Tpr hash 00/wudata_01.tpr: 3088256186 2730024186 1947096957 1214887156 2277528353
14:39:52:WU00:FS01:0x15:GPU device id=0
14:39:52:WU00:FS01:0x15:Working on Protein
14:39:52:WU00:FS01:0x15:Client config unavailable.
14:39:53:WU00:FS01:0x15:Starting GUI Server
14:40:59:WU00:FS01:0x15:Resuming from checkpoint
14:40:59:WU00:FS01:0x15:fcCheckPointResume: retreived and current tpr file hash:
14:40:59:WU00:FS01:0x15: 0 3088256186 3088256186
14:40:59:WU00:FS01:0x15: 1 2730024186 2730024186
14:40:59:WU00:FS01:0x15: 2 1947096957 1947096957
14:40:59:WU00:FS01:0x15: 3 1214887156 1214887156
14:40:59:WU00:FS01:0x15: 4 2277528353 2277528353
14:40:59:WU00:FS01:0x15:fcCheckPointResume: file hashes same.
14:40:59:WU00:FS01:0x15:fcCheckPointResume: state restored.
14:40:59:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.log Verified 00/wudata_01.log
14:40:59:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.trr Verified 00/wudata_01.trr
14:40:59:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.xtc Verified 00/wudata_01.xtc
14:40:59:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.edr Verified 00/wudata_01.edr
14:40:59:WU00:FS01:0x15:fcCheckPointResume: state restored 2
14:40:59:WU00:FS01:0x15:Resumed from checkpoint
14:40:59:WU00:FS01:0x15:Setting checkpoint frequency: 400000
14:40:59:WU00:FS01:0x15:Completed 16800001 out of 40000000 steps (42%).
14:54:12:FS01:Paused
14:54:12:FS01:Shutting core down
14:54:20:WU00:FS01:0x15:Client no longer detected. Shutting down core
14:54:20:WU00:FS01:0x15:
14:54:20:WU00:FS01:0x15:Folding@home Core Shutdown: CLIENT_DIED
14:54:21:WU00:FS01:FahCore returned: INTERRUPTED (102 = 0x66)
14:56:17:FS01:Unpaused
14:56:17:WU00:FS01:Starting
14:56:17:WU00:FS01:Running FahCore: "C:\Program Files\FAHClient/FAHCoreWrapper.exe" "C:/Documents and Settings/redacted/Application Data/FAHClient/cores/www.stanford.edu/~pande/Win32/x86/NVIDIA/Fermi/Core_15.fah/FahCore_15.exe" -dir 00 -suffix 01 -version 703 -lifeline 1880 -checkpoint 15 -gpu 0 -gpu-vendor nvidia
14:56:17:WU00:FS01:Started FahCore on PID 1888
14:56:17:WU00:FS01:Core PID:3580
14:56:17:WU00:FS01:FahCore 0x15 started
14:56:18:WU00:FS01:0x15:
14:56:18:WU00:FS01:0x15:*------------------------------*
14:56:18:WU00:FS01:0x15:Folding@Home GPU Core
14:56:18:WU00:FS01:0x15:Version 2.25 (Wed May 9 17:03:01 EDT 2012)
14:56:18:WU00:FS01:0x15:Build host AmoebaRemote
14:56:18:WU00:FS01:0x15:Board Type NVIDIA/CUDA
14:56:18:WU00:FS01:0x15:Core 15
14:56:18:WU00:FS01:0x15:
14:56:18:WU00:FS01:0x15:Window's signal control handler registered.
14:56:18:WU00:FS01:0x15:Preparing to commence simulation
14:56:18:WU00:FS01:0x15:- Looking at optimizations...
14:56:18:WU00:FS01:0x15:- Files status OK
14:56:18:WU00:FS01:0x15:sizeof(CORE_PACKET_HDR) = 512 file=<>
14:56:18:WU00:FS01:0x15:- Expanded 126190 -> 507182 (decompressed 401.9 percent)
14:56:18:WU00:FS01:0x15:Called DecompressByteArray: compressed_data_size=126190 data_size=507182, decompressed_data_size=507182 diff=0
14:56:18:WU00:FS01:0x15:- Digital signature verified
14:56:18:WU00:FS01:0x15:
14:56:18:WU00:FS01:0x15:Project: 7626 (Run 629, Clone 3, Gen 15)
14:56:18:WU00:FS01:0x15:
14:56:18:WU00:FS01:0x15:Assembly optimizations on if available.
14:56:18:WU00:FS01:0x15:Entering M.D.
14:56:20:WU00:FS01:0x15:Will resume from checkpoint file 00/wudata_01.ckp
14:56:20:WU00:FS01:0x15:Tpr hash 00/wudata_01.tpr: 3088256186 2730024186 1947096957 1214887156 2277528353
14:56:20:WU00:FS01:0x15:GPU device id=0
14:56:20:WU00:FS01:0x15:Working on Protein
14:56:20:WU00:FS01:0x15:Client config unavailable.
14:56:20:WU00:FS01:0x15:Starting GUI Server
14:57:26:WU00:FS01:0x15:Resuming from checkpoint
14:57:26:WU00:FS01:0x15:fcCheckPointResume: retreived and current tpr file hash:
14:57:26:WU00:FS01:0x15: 0 3088256186 3088256186
14:57:26:WU00:FS01:0x15: 1 2730024186 2730024186
14:57:26:WU00:FS01:0x15: 2 1947096957 1947096957
14:57:26:WU00:FS01:0x15: 3 1214887156 1214887156
14:57:26:WU00:FS01:0x15: 4 2277528353 2277528353
14:57:26:WU00:FS01:0x15:fcCheckPointResume: file hashes same.
14:57:26:WU00:FS01:0x15:fcCheckPointResume: state restored.
14:57:26:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.log Verified 00/wudata_01.log
14:57:26:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.trr Verified 00/wudata_01.trr
14:57:26:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.xtc Verified 00/wudata_01.xtc
14:57:26:WU00:FS01:0x15:fcCheckPointResume: name 00/wudata_01.edr Verified 00/wudata_01.edr
14:57:26:WU00:FS01:0x15:fcCheckPointResume: state restored 2
14:57:26:WU00:FS01:0x15:Resumed from checkpoint
14:57:26:WU00:FS01:0x15:Setting checkpoint frequency: 400000
14:57:26:WU00:FS01:0x15:Completed 16800001 out of 40000000 steps (42%).
14:57:27:WARNING:WU00:FS01:Detected clock skew (1 mins 10 secs), adjusting time estimates
15:14:28:WU00:FS01:0x15:Completed 17200000 out of 40000000 steps (43%).
15:17:23:FS01:Finishing
15:30:16:WU00:FS01:0x15:Completed 17600000 out of 40000000 steps (44%).
15:48:37:WU00:FS01:0x15:Completed 18000000 out of 40000000 steps (45%).
16:09:10:WU00:FS01:0x15:Completed 18400000 out of 40000000 steps (46%).
16:24:53:WU00:FS01:0x15:Completed 18800000 out of 40000000 steps (47%).
16:40:35:WU00:FS01:0x15:Completed 19200000 out of 40000000 steps (48%).
16:56:18:WU00:FS01:0x15:Completed 19600000 out of 40000000 steps (49%).
17:12:00:WU00:FS01:0x15:Completed 20000000 out of 40000000 steps (50%).
17:27:43:WU00:FS01:0x15:Completed 20400000 out of 40000000 steps (51%).
17:43:25:WU00:FS01:0x15:Completed 20800000 out of 40000000 steps (52%).
17:59:08:WU00:FS01:0x15:Completed 21200000 out of 40000000 steps (53%).
18:14:50:WU00:FS01:0x15:Completed 21600000 out of 40000000 steps (54%).
18:30:32:WU00:FS01:0x15:Completed 22000000 out of 40000000 steps (55%).
18:46:15:WU00:FS01:0x15:Completed 22400000 out of 40000000 steps (56%).
19:01:57:WU00:FS01:0x15:Completed 22800000 out of 40000000 steps (57%).
19:17:40:WU00:FS01:0x15:Completed 23200000 out of 40000000 steps (58%).
19:33:22:WU00:FS01:0x15:Completed 23600000 out of 40000000 steps (59%).
19:49:05:WU00:FS01:0x15:Completed 24000000 out of 40000000 steps (60%).
20:04:47:WU00:FS01:0x15:Completed 24400000 out of 40000000 steps (61%).
20:20:30:WU00:FS01:0x15:Completed 24800000 out of 40000000 steps (62%).
******************************* Date: 2013-02-26 *******************************
20:36:13:WU00:FS01:0x15:Completed 25200000 out of 40000000 steps (63%).
20:51:55:WU00:FS01:0x15:Completed 25600000 out of 40000000 steps (64%).
21:07:38:WU00:FS01:0x15:Completed 26000000 out of 40000000 steps (65%).
21:23:21:WU00:FS01:0x15:Completed 26400000 out of 40000000 steps (66%).
21:39:04:WU00:FS01:0x15:Completed 26800000 out of 40000000 steps (67%).
21:54:46:WU00:FS01:0x15:Completed 27200000 out of 40000000 steps (68%).
22:10:29:WU00:FS01:0x15:Completed 27600000 out of 40000000 steps (69%).
22:26:12:WU00:FS01:0x15:Completed 28000000 out of 40000000 steps (70%).
22:41:54:WU00:FS01:0x15:Completed 28400000 out of 40000000 steps (71%).
22:57:37:WU00:FS01:0x15:Completed 28800000 out of 40000000 steps (72%).
23:13:20:WU00:FS01:0x15:Completed 29200000 out of 40000000 steps (73%).
23:29:02:WU00:FS01:0x15:Completed 29600000 out of 40000000 steps (74%).
23:44:45:WU00:FS01:0x15:Completed 30000000 out of 40000000 steps (75%).
00:00:27:WU00:FS01:0x15:Completed 30400000 out of 40000000 steps (76%).
00:16:10:WU00:FS01:0x15:Completed 30800000 out of 40000000 steps (77%).
00:31:53:WU00:FS01:0x15:Completed 31200000 out of 40000000 steps (78%).
00:47:35:WU00:FS01:0x15:Completed 31600000 out of 40000000 steps (79%).
01:03:18:WU00:FS01:0x15:Completed 32000000 out of 40000000 steps (80%).
01:19:01:WU00:FS01:0x15:Completed 32400000 out of 40000000 steps (81%).
01:34:43:WU00:FS01:0x15:Completed 32800000 out of 40000000 steps (82%).
01:50:26:WU00:FS01:0x15:Completed 33200000 out of 40000000 steps (83%).
02:06:09:WU00:FS01:0x15:Completed 33600000 out of 40000000 steps (84%).
02:21:51:WU00:FS01:0x15:Completed 34000000 out of 40000000 steps (85%).
******************************* Date: 2013-02-27 *******************************
02:37:34:WU00:FS01:0x15:Completed 34400000 out of 40000000 steps (86%).
02:53:16:WU00:FS01:0x15:Completed 34800000 out of 40000000 steps (87%).
03:08:59:WU00:FS01:0x15:Completed 35200000 out of 40000000 steps (88%).
03:24:42:WU00:FS01:0x15:Completed 35600000 out of 40000000 steps (89%).
03:40:24:WU00:FS01:0x15:Completed 36000000 out of 40000000 steps (90%).
03:56:07:WU00:FS01:0x15:Completed 36400000 out of 40000000 steps (91%).
04:11:50:WU00:FS01:0x15:Completed 36800000 out of 40000000 steps (92%).
04:27:33:WU00:FS01:0x15:Completed 37200000 out of 40000000 steps (93%).
04:43:15:WU00:FS01:0x15:Completed 37600000 out of 40000000 steps (94%).
04:58:58:WU00:FS01:0x15:Completed 38000000 out of 40000000 steps (95%).
05:14:41:WU00:FS01:0x15:Completed 38400000 out of 40000000 steps (96%).
05:30:23:WU00:FS01:0x15:Completed 38800000 out of 40000000 steps (97%).
05:46:06:WU00:FS01:0x15:Completed 39200000 out of 40000000 steps (98%).
06:01:49:WU00:FS01:0x15:Completed 39600000 out of 40000000 steps (99%).
06:17:31:WU00:FS01:0x15:Completed 40000000 out of 40000000 steps (100%).
06:17:32:WU00:FS01:0x15:Finished fah_main status=0
06:17:32:WU00:FS01:0x15:Successful run
06:17:32:WU00:FS01:0x15:DynamicWrapper: Finished Work Unit: sleep=10000
06:17:42:WU00:FS01:0x15:Reserved 768760 bytes for xtc file; Cosm status=0
06:17:42:WU00:FS01:0x15:Allocated 768760 bytes for xtc file
06:17:42:WU00:FS01:0x15:- Reading up to 768760 from "00/wudata_01.xtc": Read 768760
06:17:42:WU00:FS01:0x15:Read 768760 bytes from xtc file; available packet space=785661704
06:17:42:WU00:FS01:0x15:xtc file hash check passed.
06:17:42:WU00:FS01:0x15:Reserved 48264 48264 785661704 bytes for arc file=<00/wudata_01.trr> Cosm status=0
06:17:42:WU00:FS01:0x15:Allocated 48264 bytes for arc file
06:17:42:WU00:FS01:0x15:- Reading up to 48264 from "00/wudata_01.trr": Read 48264
06:17:42:WU00:FS01:0x15:Read 48264 bytes from arc file; available packet space=785613440
06:17:42:WU00:FS01:0x15:trr file hash check passed.
06:17:42:WU00:FS01:0x15:Allocated 544 bytes for edr file
06:17:42:WU00:FS01:0x15:Read bedfile
06:17:42:WU00:FS01:0x15:edr file hash check passed.
06:17:42:WU00:FS01:0x15:Allocated 37154 bytes for logfile
06:17:42:WU00:FS01:0x15:Read logfile
06:17:42:WU00:FS01:0x15:GuardedRun: success in DynamicWrapper
06:17:42:WU00:FS01:0x15:GuardedRun: done
06:17:42:WU00:FS01:0x15:Run: GuardedRun completed.
06:17:43:WU00:FS01:0x15:+ Opened results file
06:17:43:WU00:FS01:0x15:- Writing 855234 bytes of core data to disk...
06:17:43:WU00:FS01:0x15:Done: 854722 -> 826358 (compressed to 96.6 percent)
06:17:43:WU00:FS01:0x15: ... Done.
06:17:43:WU00:FS01:0x15:DeleteFrameFiles: successfully deleted file=00/wudata_01.ckp
06:17:43:WU00:FS01:0x15:Shutting down core
06:17:43:WU00:FS01:0x15:
06:17:43:WU00:FS01:0x15:Folding@home Core Shutdown: FINISHED_UNIT
06:17:43:WU00:FS01:FahCore returned: FINISHED_UNIT (100 = 0x64)
06:17:43:WU00:FS01:Sending unit results: id:00 state:SEND error:NO_ERROR project:7626 run:629 clone:3 gen:15 core:0x15 unit:0x00000013664f2dd14fe61d5f78a5f4d0
06:17:43:WU00:FS01:Uploading 807.49KiB to 171.64.65.105
06:17:43:WU00:FS01:Connecting to 171.64.65.105:8080
06:17:45:WU00:FS01:Upload complete
06:17:46:WU00:FS01:Server responded WORK_ACK (400)
06:17:46:WU00:FS01:Final credit estimate, 14093.00 points
06:17:46:WU00:FS01:Cleaning up
Re: Sanity check: Is this proper Kepler throttling?
Posted: Mon Mar 04, 2013 10:49 pm
by bruce
Divide the number of steps by the total number of steps in more than one line such as these two.
15:48:37:WU00:FS01:0x15:Completed 18000000 out of 40000000 steps (45%).
18:46:15:WU00:FS01:0x15:Completed 22400000 out of 40000000 steps (56%).
In most cases, you'll get the percentage shown at the end of the line (but occasionally the percentage contains a fractional percentage that is rounded off). Subtract:
56%-45% = 11%
18:46:15-15:48:37 = 3h1m38s
Divide: 3h1m38s / 11 = 999 seconds per frame (aka percent) = 16m31s
That's an average over 11 frames. Of course it's easier to just pick one specific percent.
16:09:10:WU00:FS01:0x15:Completed 18400000 out of 40000000 steps (46%).
16:24:53:WU00:FS01:0x15:Completed 18800000 out of 40000000 steps (47%).
16:24:53-16:09:10 = 8m43s
or
06:01:49:WU00:FS01:0x15:Completed 39600000 out of 40000000 steps (99%).
06:17:31:WU00:FS01:0x15:Completed 40000000 out of 40000000 steps (100%).
15m42s
Clearly the time is changing very significantly which usually means that FAH is competing with other programs and there's no way to establish a single value that applies to each frame.
Re: Sanity check: Is this proper Kepler throttling?
Posted: Tue Mar 05, 2013 2:01 am
by JimF
coredump4,
I can give you a fairly direct comparison, with both a GTX 650 Ti (twice the shaders as a GTX 650) and a GTX 560 (non-Ti) on the same motherboard. I was averaging about the same PPD on each for Folding, depending on the project. The harder ones (more complex molecules presumably) favored the GTX 650 Ti with its larger number of shaders, while the simpler ones favored the GTX 560 with its higher shader clock rate. I don't have them running anymore, since they do not get along well for Folding, but the bottom line is that you should expect to get about 1/2 the PPD on your overclocked GTX 650 as on your GTX 560 Ti (which I also have, but on a different PC).
Re: Sanity check: Is this proper Kepler throttling?
Posted: Tue Mar 05, 2013 2:25 am
by coredump4
Bruce,
Thanks for that explanation!
JimF,
Thanks for the comparison. It sounds like my 650 is behaving properly, I just overestimated it relative to the 560 Ti. I really like the 560 Ti, and I got it as an open-box item for $77, so that was quite worth it. Not bleeding edge anymore, but still quite good for folding.