CUDA not running?

Seti just keeps downloading new workunits and none of them are running (shows as ready to start). I only allow GPU no CPU. I stopped it from requesting more task but it is still not running. I suspend my other works but it still refuses to run.

Use GPU while computer is in use: checked
While processor usage is less than 0 percent

BOINC version 6.10.36
GPU: nVidia GeForce 9600GSO (driver 196.21)

Your CUDA card probably hasn't enough available memory.

See this forum thread for more info.

Gruß,
Gundolf
____________
Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

i got this, that's all

4/04/2010 19:15:27 PM SETI@home [coproc_debug] Assigning CUDA instance 0 to 11fe07ae.20579.12751.5.10.36_1

What are the settings in the Activity menu?

Restart BOINC and post here the first 25-30 lines from Messages tab.

.
____________



- ALF - "Find out what you don't do well ..... then don't do it!" :)

5/04/2010 0:24:07 AM Starting BOINC client version 6.10.36 for windows_x86_64
5/04/2010 0:24:07 AM log flags: file_xfer, sched_ops, task
5/04/2010 0:24:07 AM Libraries: libcurl/7.19.7 OpenSSL/0.9.8l zlib/1.2.3
5/04/2010 0:24:07 AM Data directory: C:\ProgramData\BOINC
5/04/2010 0:24:07 AM Running under account RJ(ern)
5/04/2010 0:24:07 AM Processor: 4 AuthenticAMD AMD Phenom(tm) II X4 B55 Processor [Family 16 Model 4 Stepping 3]
5/04/2010 0:24:07 AM Processor: 512.00 KB cache
5/04/2010 0:24:07 AM Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 htt pni cx16 syscall nx lm svm sse4a osvw ibs skinit wdt page1gb rdtscp 3dnowext 3dnow
5/04/2010 0:24:07 AM OS: Microsoft Windows 7: Ultimate x64 Edition, (06.01.7600.00)
5/04/2010 0:24:07 AM Memory: 4.00 GB physical, 8.00 GB virtual
5/04/2010 0:24:07 AM Disk: 227.34 GB total, 175.93 GB free
5/04/2010 0:24:07 AM Local time is UTC +8 hours
5/04/2010 0:24:07 AM NVIDIA GPU 0: GeForce 9600 GSO (driver version 19621, CUDA version 3000, compute capability 1.1, 384MB, 264 GFLOPS peak)
5/04/2010 0:24:07 AM rosetta@home URL http://boinc.bakerlab.org/rosetta/; Computer ID 1263675; resource share 100
5/04/2010 0:24:07 AM Einstein@Home URL http://einstein.phys.uwm.edu/; Computer ID 2569289; resource share 100
5/04/2010 0:24:07 AM SETI@home URL http://setiathome.berkeley.edu/; Computer ID 5360182; resource share 100
5/04/2010 0:24:07 AM SETI@home General prefs: from SETI@home (last modified 04-Apr-2010 17:58:49)
5/04/2010 0:24:07 AM SETI@home Computer location: home
5/04/2010 0:24:07 AM SETI@home General prefs: no separate prefs for home; using your defaults
5/04/2010 0:24:07 AM Reading preferences override file
5/04/2010 0:24:07 AM Preferences:
5/04/2010 0:24:07 AM max memory usage when active: 818.90MB
5/04/2010 0:24:07 AM max memory usage when idle: 2866.14MB
5/04/2010 0:24:07 AM max disk usage: 3.00GB
5/04/2010 0:24:07 AM max CPUs used: 2
5/04/2010 0:24:07 AM (to change, visit the web site of an attached project,
5/04/2010 0:24:07 AM or click on Preferences)
5/04/2010 0:24:07 AM Not using a proxy
5/04/2010 0:24:08 AM Einstein@Home Restarting task h1_0436.70_S5R4__126_S5GCEa_1 using einstein_S5GCE version 304
5/04/2010 0:24:08 AM rosetta@home Restarting task lr5_dun08_slow_corrections_semirot_it02_A_rlbd_1xd6_SAVE_ALL_OUT_IGNORE_THE_REST_DECOY_18081_1998_0 using minirosetta version 205

Why didn't you switch on the <cpu_sched_debug> logging option, as described in the threads I linked to?

Gruß,
Gundolf

i have already posted that. it is the same for all WUs. I have seti stop requesting for new task for now

4/04/2010 19:15:27 PM SETI@home [coproc_debug] Assigning CUDA instance 0 to 11fe07ae.20579.12751.5.10.36_1dy replied to that

When you last rebooted your computer?
This may free GPU memory.

You can use GPU-Z to check how much GPU RAM is used/free.
http://www.techpowerup.com/gpuz/

"Use GPU while computer is in use: checked":
checked both on web site and in BOINC Manager preferences?

Try BOINC 6.10.43
http://boinc.berkeley.edu/download_all.php

Try NVIDIA driver 197.13 (faster than 196.21)
http://setiathome.berkeley.edu/forum_thread.php?id=59520

.
____________



- ALF - "Find out what you don't do well ..... then don't do it!" :)

Why didn't you switch on the <cpu_sched_debug> logging option, as described in the threads I linked to?

i have already posted that. it is the same for all WUs. I have seti stop requesting for new task for now

4/04/2010 19:15:27 PM SETI@home [coproc_debug] Assigning CUDA instance 0 to 11fe07ae.20579.12751.5.10.36_1

dy replied to that

5/04/2010 19:54:01 PM SETI@home [cpu_sched_debug] 15fe07ae.8052.119197.6.10.25_0: insufficient GPU RAM (230MB < 368MB)

Seti needs at least 254MB free video RAM for CUDA 2.1 and 2.2,
Seti needs at least 384MB free video RAM for CUDA 2.3 and above.

These minimums are set by which driver you use, BOINC will detect that and use those automatically on the scheduler contact.

What happens when you reboot? Still no changes?
Are you running the Aero desktop in Windows 7? That one sucks up video RAM. What if you disable it? What is the resolution and bit-depth you run your desktop in?
What if you lower those?

If none of that helps, you may have to return in driver version, to a lower one. All minimum driver version for using CUDA is 177.35, for CUDA 2.3 it's 190.38
____________
Jord

- BOINC FAQ Service
- BOINC User Wiki

Real is just a matter of perception.

The new NVIDIA drivers 197.13+ use less GPU memory

Did you read:
GPU RAM: any ideas on how to free up a few MB?
http://setiathome.berkeley.edu/forum_thread.php?id=59276

My post in the above thread "If you install Lunatics' Unified Installer v0.2 ...... Requires 220 MB free Video RAM"
http://setiathome.berkeley.edu/forum_thread.php?id=59276&nowrap=true#980666


____________



- ALF - "Find out what you don't do well ..... then don't do it!" :)