Advanced search

Message boards : Number crunching : Error While Computing With My New GPU

Author Message
Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44062 - Posted: 28 Jul 2016 | 15:42:41 UTC

Hello all,

Yesterday I got myself a new gtx 960 from Best Buy and I ran a project with it for 19 hours. It appears that I woke up today and the project is off of my computer, but according to the website had it came as an error.

Any way I can fix this?

Thank you.

Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44065 - Posted: 29 Jul 2016 | 21:45:20 UTC - in response to Message 44062.

Anyone, please?

Jim1348
Send message
Joined: 28 Jul 12
Posts: 819
Credit: 1,591,285,971
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 44066 - Posted: 29 Jul 2016 | 22:03:17 UTC - in response to Message 44065.

It may be overclocked too much.
http://www.gpugrid.net/forum_thread.php?id=3299#28672

Profile skgiven
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 23 Apr 09
Posts: 3968
Credit: 1,995,359,260
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 44067 - Posted: 29 Jul 2016 | 22:24:17 UTC - in response to Message 44066.
Last modified: 29 Jul 2016 | 22:27:10 UTC

1266MHz might be too high as a base clock (FOC) - reference is 1127 (base)/1178 (Boost)MHz.
You could use a tool such as MSI Afterburner to reduce the frequency.
____________
FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44068 - Posted: 29 Jul 2016 | 23:26:35 UTC - in response to Message 44067.

1266MHz might be too high as a base clock (FOC) - reference is 1127 (base)/1178 (Boost)MHz.
You could use a tool such as MSI Afterburner to reduce the frequency.



How can I underclock it when MSI afterburner says my core clock is "0"?

Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44069 - Posted: 29 Jul 2016 | 23:30:59 UTC - in response to Message 44066.

It may be overclocked too much.
http://www.gpugrid.net/forum_thread.php?id=3299#28672


Hmmm,

I see it talks about clean install drivers. Do you know what the best driver is for nvidia cards on this website, or are each different?

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 44070 - Posted: 30 Jul 2016 | 8:46:58 UTC - in response to Message 44069.
Last modified: 30 Jul 2016 | 8:52:55 UTC

359.6 are fastest for version of CUDA this project is running.

I would also lower your "additional work" buffer to 0.2 and "Store up to an additional" to 0 so that you can finish each WU in the 24hr deadline for full bonus.

Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44071 - Posted: 30 Jul 2016 | 13:37:18 UTC - in response to Message 44070.

359.6 are fastest for version of CUDA this project is running.

I would also lower your "additional work" buffer to 0.2 and "Store up to an additional" to 0 so that you can finish each WU in the 24hr deadline for full bonus.



I don't see an option where it says "additional work" but I see the 2nd option you posted and I changed that to 0

Do you mean "store up to at least "" days of work"? mine is already set up to 0.1

Thanks

Jim1348
Send message
Joined: 28 Jul 12
Posts: 819
Credit: 1,591,285,971
RAC: 0
Level
His
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 44072 - Posted: 30 Jul 2016 | 13:59:14 UTC - in response to Message 44071.
Last modified: 30 Jul 2016 | 14:05:45 UTC

There is a discussion of the driver issue, and yes it is 359.06:
http://www.gpugrid.net/forum_thread.php?id=4319

And you can obtain a zero buffer size by setting the GPUGrid resource share to zero:
https://www.gpugrid.net/forum_thread.php?id=3986#39554

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 44073 - Posted: 30 Jul 2016 | 14:01:47 UTC - in response to Message 44071.

359.6 are fastest for version of CUDA this project is running.

I would also lower your "additional work" buffer to 0.2 and "Store up to an additional" to 0 so that you can finish each WU in the 24hr deadline for full bonus.



I don't see an option where it says "additional work" but I see the 2nd option you posted and I changed that to 0

Do you mean "store up to at least "" days of work"? mine is already set up to 0.1

Thanks


"Store at least set to = 0.2
"Additional work" set to = 0

Profile Logan Carr
Send message
Joined: 12 Aug 15
Posts: 240
Credit: 64,069,811
RAC: 0
Level
Thr
Scientific publications
watwatwatwat
Message 44074 - Posted: 30 Jul 2016 | 19:24:05 UTC - in response to Message 44073.

Thank you both I greatly appreciate it. So far everything looks fine.

Post to thread

Message boards : Number crunching : Error While Computing With My New GPU

//