I'd split the difference and go to 6.2.19.
Printable View
I'd split the difference and go to 6.2.19.
What are the advantages if any of moving from 5.10.45 x64 (which has been a flawless performer) to the newer versions?
The CUDA support of 6.4.5 is about the only reason I know of. So far, I have not found a way to take advantage of it yet. For those running old cards or headless, I see no advantage to running 6.4.5. I'm still running 5.10.45 64 bit on most of the farm since they are very old graphic cards there. 5.10.45 has been bullet proof for me.
I'm having the odd issue of queuing WUs, as I reported on the other thread with high priority WUs, on the only machine running 6.4.5. I suspect this may be due to 6.4.5. I won't propagate it to any other machines until I know what's up with that...
Regards,
Bob
6.5.0 beta is available for:
Linux 32 bit
Linux 64 bit
New year, new beta, this time for OS X users:
6.6.0 beta is available for:
OS X
OS X command line version
6.6.0 beta is available for:
Windows 32 bit
Windows 64 bit
just installed v6.4.5 win32
i have a couple of qustions
1. i just have 4 active WUs and don't get anyone new for pending, what's wrong?
2. and how to switch from high priority to normal
http://localhostr.com/files/562809/boinc645.png
as u can see in shot, 3 WUs have high priority and 1 is normal
You will get new WUs as soon as one is nearly finished or as soon as it has finished. I've also got it from time to time.
yea, but i have them in previous version and now they are gone :)
one WU is done and gone for reporting, three are crunching but no new ones
1/15/2009 4:01:42 PM|World Community Grid|Sending scheduler request: Requested by user. Requesting 0 seconds of work, reporting 1 completed tasks
1/15/2009 4:01:47 PM|World Community Grid|Scheduler request completed: got 0 new tasks
still no new WUs
http://localhostr.com/files/4d9e11/boinc645.png
what could be wrong?
edit: i've just detached GPUGrid Project and recieved 2 WUs from WCG
6.6.1 beta is available for:
OS X
OS X command line version
Check your client preferences. The (retarded) defaults are to connect every 4 days and ask for 1 day's worth of work. Make sure you're connecting at least daily and asking for at least a day's worth of work. Then make sure your network activity isn't restricted somehow, either in the "activity" menu or in your preferences.
this is how it's set
http://localhostr.com/files/149ed5/boinc.png
i get WU at the same moment that there is one finished. always there are 4 active WUs
but in 5.10.45 i have a lot more WUs , besides those 4 active that are waiting for crunching. now in tasks window i have only 4 WUs (those active ones)
additional work buffer= extra cache,set appropriately.
Try these settings:
Attachment 93045
back to 5.10.45...
faah4976_ 005097_ MC_ xMut_ md17820_ 02_ 0-- Inconclusive
<core_client_version>6.4.5</core_client_version>
<![CDATA[
<stderr_txt>
Failed to get VersionInfo size: 1812
INFO:[11:54:51] Start AutoGrid...
6.6.2 beta is available for:
Windows 32 bit
Windows 64 bit
Installing it now.
Interesting...
01/21/09 11:13:16|World Community Grid|Sending scheduler request: To fetch work.
01/21/09 11:13:16|World Community Grid|CPU work request: 0.00 seconds, 0 instances
01/21/09 11:13:16|World Community Grid|CUDA work request: 1036800.00 seconds, 2 instances
My benches seem to have dropped a bit, though.
CUDA for WCG? I'm on ATI so I didn't received that message. :confused:
BTW it is interesting.
I've installed Windows 7 64 bit a few days ago.
Vista x86:
2678 FP MIPS
4883 INT MIPS
Windows 7 x64:
2560 FP MIPS
6474 INT MIPS
Update:
6.6.2 beta is available for:
OS X
OS X command line version
6.6.2 beta is available for:
Linux 32 bit
Linux 64 bit