Since the last outage...

Message boards : Number crunching : Since the last outage...
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1239590 - Posted: 1 Jun 2012, 19:07:07 UTC - in response to Message 1239493.  
Last modified: 1 Jun 2012, 19:32:52 UTC

And now this:

6/1/2012 9:10:28 AM | SETI@home | update requested by user
6/1/2012 9:10:29 AM | SETI@home | Sending scheduler request: Requested by user.
6/1/2012 9:10:29 AM | SETI@home | Reporting 1 completed task, requesting new tasks for CPU and NVIDIA
6/1/2012 9:10:37 AM | SETI@home | Scheduler request completed: got 0 new tasks
6/1/2012 9:10:37 AM | SETI@home | Project has no tasks available

/me shakes his head in disappointment

32 CPU cores and 300+ GPU cores on this machine and all I can do now is one GPU workload at a time...sad.

What cache settings are you using? Don't try and use a too big cache setting as the Server's scheduler will fill your fastest resource first,
the other thing to note is that Boinc 7 uses a totally rewritten scheduler, and the Cache settings suitable for it are different to Boinc 6,
for Boinc 7 use something like:

Maintain enough tasks to keep busy for at least 2 days
(max 10 days).

... and up to an additional 0.01 days

This will cache 2.01 days work maximum, once the cache drops down to below 2 days work, Boinc will ask for work again,

Claggy
ID: 1239590 · Report as offensive
Mr. GoodWrench

Send message
Joined: 26 Jun 99
Posts: 19
Credit: 8,937,626
RAC: 0
United States
Message 1239614 - Posted: 1 Jun 2012, 20:01:58 UTC - in response to Message 1238744.  

Thanks much. Went back to an older driver and everything is working again.
ID: 1239614 · Report as offensive
Profile BilBg
Volunteer tester
Avatar

Send message
Joined: 27 May 07
Posts: 3720
Credit: 9,385,827
RAC: 0
Bulgaria
Message 1240917 - Posted: 3 Jun 2012, 23:39:02 UTC - in response to Message 1239614.  


driver: 295.73 have "sleep bug" if your monitor is connected to DVI and you let Windows to sleep the monitor (CUDA will disappear)

(If monitor is connected to VGA - No problem,
If monitor do not auto-sleep (even on DVI) - No problem)


 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 
ID: 1240917 · Report as offensive
Profile gcpeters
Avatar

Send message
Joined: 20 May 99
Posts: 67
Credit: 109,352,237
RAC: 1
United States
Message 1241097 - Posted: 4 Jun 2012, 15:24:15 UTC

Nothing works so far. I went back to version 6. Still same problem. All I get are cuda_fermi tasks.

Grrr...
ID: 1241097 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22555
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1241101 - Posted: 4 Jun 2012, 15:33:26 UTC

Until your cache is filled with enough GPU tasks that's all you'll get. There are various tricks, like turning off GPU processing until you've got enough CPU tasks that work sometimes. Of course non of them work if you are running other (CPU only) projects and they have filled the CPU part of the cache...
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1241101 · Report as offensive
Profile BilBg
Volunteer tester
Avatar

Send message
Joined: 27 May 07
Posts: 3720
Credit: 9,385,827
RAC: 0
Bulgaria
Message 1241282 - Posted: 4 Jun 2012, 19:59:45 UTC - in response to Message 1241097.  
Last modified: 4 Jun 2012, 20:06:58 UTC

I went back to version 6

"version 6" of what?

What do you see now about VRAM available?:
5/28/2012 11:49:53 AM | | NVIDIA GPU 0: GeForce 8600M GT (driver version 301.42, CUDA version 4.20, compute capability 1.1, 256MB, 176MB available, 91 GFLOPS peak)
5/28/2012 11:49:53 AM | | OpenCL: NVIDIA GPU 0: GeForce 8600M GT (driver version 301.42, device version OpenCL 1.0 CUDA, 256MB, 176MB available)


And you have "sleep bug" here:
http://setiathome.berkeley.edu/results.php?hostid=5401439&offset=0&show_names=0&state=5&appid=


 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 
ID: 1241282 · Report as offensive
Previous · 1 · 2

Message boards : Number crunching : Since the last outage...


 
©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.