Questions and Answers :
GPU applications :
GTX295 woes..... somebody help me get on track
Message board moderation
Author | Message |
---|---|
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
OK, I have a gtx295 and was running it with the latest 197.13 drivers and stuff was going slow. I went ahead and did the whole driver cleaner and reverted to the 191.07 and rebooted. With SLI on, both GPU's are recognized but it's slow as heck. With SLI off, only one GPU is recognized but its quick with the computation. How can I get the 2nd GPU recognized with SLI off? I plugged in my TV on the other connection and extended my desktop with SLI off. Still a no-go. I want to run it with SLI off since one GPU crunches faster than both with SLI on. I'd like to have both GPU's doing work. Having trouble getting my start message copied... any ideas so I can post? OH, here is my cc_config... just copy'd pasted from one that was posted. Not sure if I screwed something up here either. - <cc_config> - <log_flags> <coproc_debug>1</coproc_debug> </log_flags> - <options> <use_all_gpus>1</use_all_gpus> </options> </cc_config> |
Gundolf Jahn Send message Joined: 19 Sep 00 Posts: 3184 Credit: 446,358 RAC: 0 |
Having trouble getting my start message copied... any ideas so I can post? Just highlight the desired line(s) (shift click) and use the "Copy selected messages" button. Gruß, Gundolf Computer sind nicht alles im Leben. (Kleiner Scherz) SETI@home classic workunits 3,758 SETI@home classic CPU time 66,520 hours |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
Ha, I didn't see the button on the left! I was trying to Cntrl-C to copy. 3/31/2010 2:23:05 PM Starting BOINC client version 6.10.18 for windows_x86_64 3/31/2010 2:23:05 PM log flags: file_xfer, sched_ops, task 3/31/2010 2:23:05 PM Libraries: libcurl/7.19.4 OpenSSL/0.9.8l zlib/1.2.3 3/31/2010 2:23:05 PM Data directory: C:\ProgramData\BOINC 3/31/2010 2:23:05 PM Running under account Sean 3/31/2010 2:23:05 PM Processor: 4 GenuineIntel Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz [Intel64 Family 6 Model 30 Stepping 5] 3/31/2010 2:23:05 PM Processor: 256.00 KB cache 3/31/2010 2:23:05 PM Processor features: fpu tsc pae nx sse sse2 pni mmx 3/31/2010 2:23:05 PM OS: Microsoft Windows 7: Home Premium x64 Edition, (06.01.7600.00) 3/31/2010 2:23:05 PM Memory: 3.99 GB physical, 7.98 GB virtual 3/31/2010 2:23:05 PM Disk: 931.41 GB total, 818.67 GB free 3/31/2010 2:23:05 PM Local time is UTC -4 hours 3/31/2010 2:23:05 PM NVIDIA GPU 0: GeForce GTX 295 (driver version 19107, CUDA version 2030, compute capability 1.3, 896MB, 709 GFLOPS peak) 3/31/2010 2:23:05 PM SETI@home Found app_info.xml; using anonymous platform 3/31/2010 2:23:05 PM Not using a proxy 3/31/2010 2:23:05 PM SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5326435; resource share 100 3/31/2010 2:23:05 PM SETI@home General prefs: from SETI@home (last modified 05-Jul-2005 06:04:48) 3/31/2010 2:23:05 PM SETI@home Computer location: home 3/31/2010 2:23:05 PM SETI@home General prefs: no separate prefs for home; using your defaults 3/31/2010 2:23:05 PM Reading preferences override file 3/31/2010 2:23:05 PM Preferences limit memory usage when active to 2043.52MB 3/31/2010 2:23:05 PM Preferences limit memory usage when idle to 3678.34MB 3/31/2010 2:23:05 PM Preferences limit disk usage to 10.00GB 3/31/2010 2:23:05 PM SETI@home Restarting task 27ja07ad.7418.18886.12.10.188_3 using setiathome_enhanced version 603 3/31/2010 2:23:05 PM SETI@home Restarting task 27ja07ad.12753.23794.14.10.218_4 using setiathome_enhanced version 603 3/31/2010 2:23:05 PM SETI@home Restarting task 09mr07ag.14205.202114.3.10.101_2 using setiathome_enhanced version 603 3/31/2010 2:23:05 PM SETI@home Restarting task 02mr07ae.18241.14387.15.10.123_2 using setiathome_enhanced version 603 3/31/2010 2:23:05 PM SETI@home Restarting task 18no09ae.14296.56694.12.10.10_0 using setiathome_enhanced version 608 |
skildude Send message Joined: 4 Oct 00 Posts: 9541 Credit: 50,759,529 RAC: 60 |
you may need to install a dumby plug. If you aren't using one please refer to google for a search for "dumby plugs" there are several really good sites that describe how to make one In a rich man's house there is no place to spit but his face. Diogenes Of Sinope |
Gundolf Jahn Send message Joined: 19 Sep 00 Posts: 3184 Credit: 446,358 RAC: 0 |
Your cc_config.xml isn't at the correct place. It has to be in C:\ProgramData\BOINC If it is seen by BOINC you'll see the "coproc_debug" flag in the "log flags" line and a confirmation message about using all GPUs. Gruß, Gundolf Computer sind nicht alles im Leben. (Kleiner Scherz) SETI@home classic workunits 3,758 SETI@home classic CPU time 66,520 hours |
FiveHamlet Send message Joined: 5 Oct 99 Posts: 783 Credit: 32,638,578 RAC: 0 |
Uninstall the 197.13 driver and install the 19107 driver it is much quicker and better. Dave |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
Your cc_config.xml isn't at the correct place. It has to be in C:\ProgramData\BOINC Hmm.... It's currently located in C:\Program Files (x86)\BOINC I'll put it in that place and get back |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
Your cc_config.xml isn't at the correct place. It has to be in C:\ProgramData\BOINC You do have a C:\programdata\boinc but it is hidden by Windoze. You can see it in the lines you copied and pasted. F. |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
Yes I do.... sorry. I didnt' realize it was hidden. Thanks. I put it in the right place and rebooted the computer, made sure my desktop was extended onto the 2nd monitor and SLI disabled. Still no luck.... 3/31/2010 3:39:39 PM Starting BOINC client version 6.10.18 for windows_x86_64 3/31/2010 3:39:39 PM Config: use all coprocessors 3/31/2010 3:39:39 PM log flags: file_xfer, sched_ops, task, coproc_debug 3/31/2010 3:39:39 PM Libraries: libcurl/7.19.4 OpenSSL/0.9.8l zlib/1.2.3 3/31/2010 3:39:39 PM Data directory: C:\ProgramData\BOINC 3/31/2010 3:39:39 PM Running under account Sean 3/31/2010 3:39:39 PM Processor: 4 GenuineIntel Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz [Intel64 Family 6 Model 30 Stepping 5] 3/31/2010 3:39:39 PM Processor: 256.00 KB cache 3/31/2010 3:39:39 PM Processor features: fpu tsc pae nx sse sse2 pni mmx 3/31/2010 3:39:39 PM OS: Microsoft Windows 7: Home Premium x64 Edition, (06.01.7600.00) 3/31/2010 3:39:39 PM Memory: 3.99 GB physical, 7.98 GB virtual 3/31/2010 3:39:39 PM Disk: 931.41 GB total, 818.53 GB free 3/31/2010 3:39:39 PM Local time is UTC -4 hours 3/31/2010 3:39:39 PM NVIDIA GPU 0: GeForce GTX 295 (driver version 19107, CUDA version 2030, compute capability 1.3, 896MB, 596 GFLOPS peak) 3/31/2010 3:39:39 PM No ATI library found. 3/31/2010 3:39:39 PM SETI@home Found app_info.xml; using anonymous platform 3/31/2010 3:39:39 PM Not using a proxy 3/31/2010 3:39:39 PM SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5326435; resource share 100 3/31/2010 3:39:39 PM SETI@home General prefs: from SETI@home (last modified 05-Jul-2005 06:04:48) 3/31/2010 3:39:39 PM SETI@home Computer location: home 3/31/2010 3:39:39 PM SETI@home General prefs: no separate prefs for home; using your defaults 3/31/2010 3:39:39 PM Reading preferences override file 3/31/2010 3:39:39 PM Preferences limit memory usage when active to 2043.52MB 3/31/2010 3:39:39 PM Preferences limit memory usage when idle to 3678.34MB 3/31/2010 3:39:39 PM Preferences limit disk usage to 10.00GB 3/31/2010 3:39:39 PM SETI@home [coproc_debug] Assigning CUDA instance 0 to 09mr07ae.497.2117.12.10.137_3 3/31/2010 3:39:40 PM SETI@home Restarting task 02mr07ae.18241.14387.15.10.123_2 using setiathome_enhanced version 603 3/31/2010 3:39:40 PM SETI@home Restarting task ap_02mr07ae_B6_P0_00315_20100331_08368.wu_1 using astropulse_v505 version 505 3/31/2010 3:39:40 PM SETI@home Restarting task 11ja07ae.11469.1299.7.10.205_1 using setiathome_enhanced version 603 3/31/2010 3:39:40 PM SETI@home Restarting task 18no09ae.14296.59148.12.10.243_0 using setiathome_enhanced version 603 3/31/2010 3:39:41 PM SETI@home Restarting task 09mr07ae.497.2117.12.10.137_3 using setiathome_enhanced version 608 3/31/2010 3:39:41 PM SETI@home Sending scheduler request: To fetch work. 3/31/2010 3:39:41 PM SETI@home Requesting new tasks for CPU 3/31/2010 3:39:45 PM SETI@home Scheduler request completed: got 0 new tasks 3/31/2010 3:39:45 PM SETI@home Message from server: (Project has no jobs available) |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
On the subject. If I enable SLI, BOINC recognizes both GPU's and runs computations on both. Here is what it looks like with SLI enabled. 3/31/2010 4:01:05 PM Starting BOINC client version 6.10.18 for windows_x86_64 3/31/2010 4:01:05 PM Config: use all coprocessors 3/31/2010 4:01:05 PM log flags: file_xfer, sched_ops, task, coproc_debug 3/31/2010 4:01:05 PM Libraries: libcurl/7.19.4 OpenSSL/0.9.8l zlib/1.2.3 3/31/2010 4:01:05 PM Data directory: C:\ProgramData\BOINC 3/31/2010 4:01:05 PM Running under account Sean 3/31/2010 4:01:05 PM Processor: 4 GenuineIntel Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz [Intel64 Family 6 Model 30 Stepping 5] 3/31/2010 4:01:05 PM Processor: 256.00 KB cache 3/31/2010 4:01:05 PM Processor features: fpu tsc pae nx sse sse2 pni mmx 3/31/2010 4:01:05 PM OS: Microsoft Windows 7: Home Premium x64 Edition, (06.01.7600.00) 3/31/2010 4:01:05 PM Memory: 3.99 GB physical, 7.98 GB virtual 3/31/2010 4:01:05 PM Disk: 931.41 GB total, 818.30 GB free 3/31/2010 4:01:05 PM Local time is UTC -4 hours 3/31/2010 4:01:05 PM NVIDIA GPU 0: GeForce GTX 295 (driver version 19107, CUDA version 2030, compute capability 1.3, 896MB, 596 GFLOPS peak) 3/31/2010 4:01:05 PM NVIDIA GPU 1: GeForce GTX 295 (driver version 19107, CUDA version 2030, compute capability 1.3, 896MB, 596 GFLOPS peak) 3/31/2010 4:01:05 PM No ATI library found. 3/31/2010 4:01:05 PM SETI@home Found app_info.xml; using anonymous platform 3/31/2010 4:01:05 PM Not using a proxy 3/31/2010 4:01:05 PM SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5326435; resource share 100 3/31/2010 4:01:05 PM SETI@home General prefs: from SETI@home (last modified 05-Jul-2005 06:04:48) 3/31/2010 4:01:05 PM SETI@home Computer location: home 3/31/2010 4:01:05 PM SETI@home General prefs: no separate prefs for home; using your defaults 3/31/2010 4:01:05 PM Reading preferences override file 3/31/2010 4:01:05 PM Preferences limit memory usage when active to 2043.52MB 3/31/2010 4:01:05 PM Preferences limit memory usage when idle to 3678.34MB 3/31/2010 4:01:05 PM Preferences limit disk usage to 10.00GB 3/31/2010 4:01:05 PM SETI@home [coproc_debug] Assigning CUDA instance 0 to 18no09ae.14296.69782.12.10.184_1 3/31/2010 4:01:05 PM SETI@home [coproc_debug] Assigning CUDA instance 1 to 30ja07ae.26292.11524.7.10.211_0 3/31/2010 4:01:05 PM SETI@home Restarting task |
Gundolf Jahn Send message Joined: 19 Sep 00 Posts: 3184 Credit: 446,358 RAC: 0 |
With SLI on, both GPU's are recognized but it's slow as heck. It shouldn't depend on SLI (any longer). Is it possible that one of the GPUs is defect? Gruß, Gundolf |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
Well, with SLI on, it is recognizing both GPU's, using both GPU's and seems to be doing ok. I have been judging if its working well or not by whether each second changes the 0.1 spot, the 0.01 spot or 0.001 spot. Right now its doing approx 0.1% each second on each GPU. I was trying to get it to work with SLI off because all over these forums people recommend SLI being off for better processing. Only problem is that with SLI off I only get one GPU to work. So, you're saying it shouldn't matter if SLI is on or off? |
Fred W Send message Joined: 13 Jun 99 Posts: 2524 Credit: 11,954,210 RAC: 0 |
Well, with SLI on, it is recognizing both GPU's, using both GPU's and seems to be doing ok. I have been judging if its working well or not by whether each second changes the 0.1 spot, the 0.01 spot or 0.001 spot. Right now its doing approx 0.1% each second on each GPU. Correct. SLI on/off does not matter with the driver version that you are using. If it is working with SLI enabled, I would just leave it there. Note that will be (sometimes large) variation in crunch-times depending in the Angle Range of particular Work Units. F. |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
OK, thanks, I'll leave SLI on then and see where that gets me. I appreciate your help/time! |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
So, does the units that you earn change if it takes twice the time to process? |
skildude Send message Joined: 4 Oct 00 Posts: 9541 Credit: 50,759,529 RAC: 60 |
that all depends. CUDA is bad at processing VLAR WU's which will have a much higher completion time but you wont get any extra credit. but a VHAR will increase in value vs. length of time to complete. again it really all depends on the angle range of the WU's In a rich man's house there is no place to spit but his face. Diogenes Of Sinope |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
Strange - sometimes it sees 3 GPUs http://setiathome.berkeley.edu/result.php?resultid=1561318977 setiathome_CUDA: Found 3 CUDA device(s): Device 1 : GeForce GTX 295 totalGlobalMem = 939524096 sharedMemPerBlock = 16384 regsPerBlock = 16384 warpSize = 32 memPitch = 262144 maxThreadsPerBlock = 512 clockRate = 1242000 totalConstMem = 65536 major = 1 minor = 3 textureAlignment = 256 deviceOverlap = 1 multiProcessorCount = 30 Device 2 : GeForce GTX 295 totalGlobalMem = 939524096 sharedMemPerBlock = 16384 regsPerBlock = 16384 warpSize = 32 memPitch = 262144 maxThreadsPerBlock = 512 clockRate = 1242000 totalConstMem = 65536 major = 1 minor = 3 textureAlignment = 256 deviceOverlap = 1 multiProcessorCount = 30 Device 3 : GeForce GTX 295 totalGlobalMem = 939524096 sharedMemPerBlock = 16384 regsPerBlock = 16384 warpSize = 32 memPitch = 262144 maxThreadsPerBlock = 512 clockRate = 1242000 totalConstMem = 65536 major = 1 minor = 3 textureAlignment = 256 deviceOverlap = 1 multiProcessorCount = 30 setiathome_CUDA: CUDA Device 2 specified, checking... Device 2: GeForce GTX 295 is okay SETI@home using CUDA accelerated device GeForce GTX 295 V12 modification by Raistmer Priority of worker thread rised successfully Priority of process adjusted successfully Total GPU memory 939524096 free GPU memory 765194240 setiathome_enhanced 6.02 Visual Studio/Microsoft C++ Build features: Non-graphics CUDA VLAR autokill enabled FFTW USE_SSE x86 CPUID: Intel(R) Core(TM) i7 CPU 860 @ 2.80GHz Cache: L1=64K L2=256K CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 libboinc: 6.3.22 Work Unit Info: ............... WU true angle range is : 2.623536 After app init: total GPU memory 939524096 free GPU memory 671870976 And here it sees 1, 2, 3 GPUs restarting the same WU http://setiathome.berkeley.edu/result.php?resultid=1561182931 __________________________ You can try ReSchedule 1.9 to move "slow" VLARs to the CPU (and not killing them) ReSchedule 1.9 (my weird direct link) http://setiathome.berkeley.edu/forum_thread.php?id=59291 . Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
Strange - sometimes it sees 3 GPUs Is that my info? I actually have another gtx295 I was trying to get going. But with the second card I get errors, dont know why. |
johall Send message Joined: 12 Feb 02 Posts: 34 Credit: 13,814,461 RAC: 0 |
I didn't want to bring the whole "my win 7 system only sees 3/4 GPU" thing into this thread, it seems to be a non-fixable problem unless you're willing to go to win xp 64bit. |
BilBg Send message Joined: 27 May 07 Posts: 3720 Credit: 9,385,827 RAC: 0 |
Strange - sometimes it sees 3 GPUs Yes - click on the above link to the result then - click on Computer ID 5326435 and see: Owner johall "But with the second card I get errors": Maybe Power supply is not Powerful enough to Power 2 cards = 4 GPUs Try another arrangement of power cables. Or swap the slots (move card-1 to slot-2 and card-2 to slot-1) . Â - ALF - "Find out what you don't do well ..... then don't do it!" :) Â |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.