Hmmmm. It looks like I won't get any work until tomorrow for E@H since the log shows that I've already processed 8 tasks for the day. How that amount is figured, I can't understand since as far as I can tell I've processed only 4 tasks for today (2 complete, 2 aborted).
Michael
Belay that last remark. Lots of tasks just showed up in my queue. Then what does that log entry mean?
You have a maximum daily quota of 40 per CPU core per day which decreases by 1 for every result that errors out or is aborted. Take a look at your list of tasks and you can see how many are erroring out or are being aborted. I'm guessing that you had reached a vastly reduced daily limit and then a couple of successful CPU tasks were reported. You can see them in the list. This would have temporarily increased your limit (each good task doubles the reduced limit up to a maximum of 40/day) which has allowed you to get more tasks. Because you have continued trashing tasks, your limit has been reduced again and it is now showing as just 1/day.
You really need to stop GPU tasks until you sort out the problem. At least a few good CPU tasks will then quickly restore your daily limit. Each good task doubles the limit so you only need 6 good tasks to go 1 -> 2 -> 4 -> 8 -> 16 -> 32 -> 40.
Thanks Gary. I've stopped all GPU tasks in my prefs. I'm running 2 GW tasks now, with 2 in the wings.
I need too look at the BOINC prefs to see if there's anything I need to setup manually to successfully process GPU tasks. I guess I should've given up by now and just process CPU tasks, but GPU worked before, so I know I can do it.
Several days of successfully processing PrimeGrid WU's with my CUDA setup shows me that the CUDA package is not my problem. Is it possible that the E@H WU's are now requiring a larger chunk of my GPU memory (I have 512MB)? What are the GPU requirements for E@H?
After continual success with running PrimeGrid GPU WU's, I turned on the flow for BRP WU's, and I'm happy to report that I have successfully processed my first E@H GPU WU since I started having problems.
I'm using an older version of CUDA (4.0.21), but I'm not going to tempt fate and upgrade to the current version. After reading about the problems with the current Windows CUDA from Nvidia, I'm not going to rock the apple cart unless absolutely necessary.
RE: RE: Hmmmm. It looks
)
You have a maximum daily quota of 40 per CPU core per day which decreases by 1 for every result that errors out or is aborted. Take a look at your list of tasks and you can see how many are erroring out or are being aborted. I'm guessing that you had reached a vastly reduced daily limit and then a couple of successful CPU tasks were reported. You can see them in the list. This would have temporarily increased your limit (each good task doubles the reduced limit up to a maximum of 40/day) which has allowed you to get more tasks. Because you have continued trashing tasks, your limit has been reduced again and it is now showing as just 1/day.
You really need to stop GPU tasks until you sort out the problem. At least a few good CPU tasks will then quickly restore your daily limit. Each good task doubles the limit so you only need 6 good tasks to go 1 -> 2 -> 4 -> 8 -> 16 -> 32 -> 40.
Cheers,
Gary.
Thanks Gary. I've stopped all
)
Thanks Gary. I've stopped all GPU tasks in my prefs. I'm running 2 GW tasks now, with 2 in the wings.
I need too look at the BOINC prefs to see if there's anything I need to setup manually to successfully process GPU tasks. I guess I should've given up by now and just process CPU tasks, but GPU worked before, so I know I can do it.
Michael
____________________________________
I've successfully run several
)
I've successfully run several WU's and now E@H is giving me plenty of work. Now back to my GPU problem.
____________________________________
Several days of successfully
)
Several days of successfully processing PrimeGrid WU's with my CUDA setup shows me that the CUDA package is not my problem. Is it possible that the E@H WU's are now requiring a larger chunk of my GPU memory (I have 512MB)? What are the GPU requirements for E@H?
____________________________________
After continual success with
)
After continual success with running PrimeGrid GPU WU's, I turned on the flow for BRP WU's, and I'm happy to report that I have successfully processed my first E@H GPU WU since I started having problems.
I'm using an older version of CUDA (4.0.21), but I'm not going to tempt fate and upgrade to the current version. After reading about the problems with the current Windows CUDA from Nvidia, I'm not going to rock the apple cart unless absolutely necessary.
____________________________________