1) Message boards : Number crunching : BSOD BCCode 116? (Message 19540)
Posted 4901 days ago by tng*
All are running stock clocks.

The 260 may be getting work, but the 470 still isn't even after a driver rollback.
2) Message boards : Number crunching : BSOD BCCode 116? (Message 19521)
Posted 4902 days ago by tng*
I have just had a third system bluescreen with a BCCode of 116, all within the last 24 hours. Is anybody else seeing problems like this? These systems are GPU crunching for GPUGRID, and two of them do nothing but crunch. The affected systems are this one (where I updated the drivers and now don't seem to be able to get work), this one, and this one.

These systems have been stable for a long time (stable by my standards -- if a system bluescreens once a month when I'm not making changes to it or pushing the envelope in some way, I'll scrap it if I can't fix it). Is anybody else seeing problems?
3) Message boards : Number crunching : Updated drivers to 260.99 -- can't get work? (Message 19520)
Posted 4902 days ago by tng*
I had a BSOD today with BCCode 116 on this system, and grabbed the latest (260.99) drivers from Nvidia. Now it doesn't seem to be getting new work. Messages at boot:

11/16/2010 7:16:18 PM Starting BOINC client version 6.10.18 for windows_x86_64
11/16/2010 7:16:18 PM log flags: file_xfer, sched_ops, task, sched_op_debug
11/16/2010 7:16:18 PM Libraries: libcurl/7.19.4 OpenSSL/0.9.8l zlib/1.2.3
11/16/2010 7:16:18 PM Data directory: C:\ProgramData\BOINC
11/16/2010 7:16:18 PM Running under account tng
11/16/2010 7:16:18 PM Processor: 16 GenuineIntel Intel(R) Xeon(R) CPU E5520 @ 2.27GHz [Intel64 Family 6 Model 26 Stepping 5]
11/16/2010 7:16:18 PM Processor: 256.00 KB cache
11/16/2010 7:16:18 PM Processor features: fpu tsc pae nx sse sse2 pni
11/16/2010 7:16:18 PM OS: Microsoft Windows Server 2008: Enterprise x64 Edition, Service Pack 2, (06.00.6002.00)
11/16/2010 7:16:18 PM Memory: 12.00 GB physical, 23.92 GB virtual
11/16/2010 7:16:18 PM Disk: 465.69 GB total, 329.78 GB free
11/16/2010 7:16:18 PM Local time is UTC -6 hours
11/16/2010 7:16:18 PM NVIDIA GPU 0: GeForce GTX 260 (driver version 26099, CUDA version 3020, compute capability 1.3, 869MB, 537 GFLOPS peak)
11/16/2010 7:16:18 PM Not using a proxy

Messages when requesting work:

11/16/2010 7:56:23 PM GPUGRID update requested by user
11/16/2010 7:56:25 PM GPUGRID [sched_op_debug] Starting scheduler request
11/16/2010 7:56:25 PM GPUGRID Sending scheduler request: Requested by user.
11/16/2010 7:56:25 PM GPUGRID Requesting new tasks for GPU
11/16/2010 7:56:25 PM GPUGRID [sched_op_debug] CPU work request: 0.00 seconds; 0 idle CPUs
11/16/2010 7:56:25 PM GPUGRID [sched_op_debug] NVIDIA GPU work request: 6637.26 seconds; 0 idle GPUs
11/16/2010 7:56:30 PM GPUGRID Scheduler request completed: got 0 new tasks
11/16/2010 7:56:30 PM GPUGRID [sched_op_debug] Server version 611
11/16/2010 7:56:30 PM GPUGRID Message from server: No work sent
11/16/2010 7:56:30 PM GPUGRID Message from server: ACEMD beta version is not available for your type of computer.
11/16/2010 7:56:30 PM GPUGRID Project requested delay of 31 seconds
11/16/2010 7:56:30 PM GPUGRID [sched_op_debug] Deferring communication for 31 sec
11/16/2010 7:56:30 PM GPUGRID [sched_op_debug] Reason: requested by project

Do I need to roll back? Or what do I need to do?
4) Message boards : Graphics cards (GPUs) : GPUGRID and Fermi (Message 16764)
Posted 5100 days ago by tng*
tng, I just put the Fermi to have a go at your suggestion.

Boinc,
03/05/2010 02:15:24 NVIDIA GPU 0: (driver version 19621, CUDA version 3000, compute capability 2.0, 1280MB, 50 GFLOPS peak) :)

Installed driver,
03/05/2010 02:40:34 NVIDIA GPU 0: GeForce GTX 470 (driver version 19741, CUDA version 3000, compute capability 2.0, 1248MB, 726 GFLOPS peak)

Set option to max performance.

Ran a Fermi Beta,
Still taking about 11min.
3 May 2010 1:37:41 UTC 3 May 2010 1:51:13 UTC Completed and validated 656.59 649.57 187.28 280.92 ACEMD beta version v6.23 (cuda30)
Boinc still reading sleeping clock rates?
http://www.gpugrid.net/result.php?resultid=2262368


Clock rate of .81 GHz is standard for a GTX 470. I don't think BOINC reads the clock rate from the card, I think it just IDs the card and reports the clock rate based on that.
5) Message boards : Graphics cards (GPUs) : GPUGRID and Fermi (Message 16759)
Posted 5100 days ago by tng*
I have the same problem with my GTX470. I had to overclock it to get the time down from 1091sec (18min) to about 690sec (11.5min). But this is a bad fix!
MrS suggested my card was defaulting to some power saving mode and the clocks were not going back up correctly. Don’t know if the problem is to do with the drivers or the card, or what, but it is not the fault of Boinc, according to MrS.

Paul, the CPU usage can speed up the project. Hence a card on a board with a 1.6GHz AMD Athlon will not get through as many tasks as a card on a board with an i7-980X.
Also, Vista/W7 slows down the project compaired to XP and Linux is much faster as it appropriates a full core to support the GPUGrid task.


Something that I ran across somewhere: this system started out doing beta WUs in ~1200 seconds. There's a setting in the NVIDIA control panel, under 3D Settings>Manage 3D settings, called "Power management mode". Setting this to "Prefer maximum performance" instead of the default "Adaptive" brought times down to below 1000 seconds. SWAN_SYNC=0 brought times down below 900 seconds, mostly around or below 850.

On the principle of "What's worth doing, is worth overdoing.", after seeing the improvement from SWAN_SYNC=0, I tried disabling hyperthreading. The one task that I ran with hyperthreading off did run faster, and with a shorter time per step, but not so much so that I was willing to continue the experiment.

[Question for project staff] Is SWAN_SYNC=0 just continuous polling?
6) Message boards : Graphics cards (GPUs) : Fermi GTX 480 all WU's fail. (Message 16412)
Posted 5114 days ago by tng*
Do you mean BOINC displays a higher GFLOPS number now? That was jsut correcting a mistake in the calculation of that rather meaningless number. Actual crunching speed shouldn't be affected by BOINC versions.


Actual runtime (12 vs. 14 minutes). I would assume that BOINC is now reporting additional capabilities that Collatz is using.
7) Message boards : Graphics cards (GPUs) : Fermi GTX 480 all WU's fail. (Message 16357)
Posted 5115 days ago by tng*
Ah -- 6.10.45. Maybe I need to see my optometrist. Significantly faster now on Collatz -- when I run through that work, I'll see if I can pull some of the GPUGRID betas.

Thanks.
8) Message boards : Graphics cards (GPUs) : Fermi GTX 480 all WU's fail. (Message 16345)
Posted 5116 days ago by tng*
Just installed a GTX 470, and I'm seeing all tasks error out immediately as well. Driver version is 197.41. Any ideas?

Reported error:

<core_client_version>6.10.43</core_client_version>
<![CDATA[
<message>
- exit code -40 (0xffffffd8)
</message>
<stderr_txt>
# There is 1 device supporting CUDA
# Device 0: "GeForce GTX 470"
# Clock rate: 0.81 GHz
# Total amount of global memory: 1309081600 bytes
# Number of multiprocessors: 14
# Number of cores: 112
SWAN : Module load result [.fastfill.cu.] [300]
SWAN: FATAL : Module load failed


</stderr_txt>
]]>



//