Advanced search

Message boards : Number crunching : Long runs yielding less credits ?

Author Message
BelgianEnthousiast
Send message
Joined: 7 Apr 15
Posts: 33
Credit: 1,201,157,375
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwat
Message 42191 - Posted: 20 Nov 2015 | 7:06:09 UTC

Hi,
For the past 10-15 days, the long runs don't bring as many credits anymore
as they used to.
Long runs since april this year yielded about 255.000 credits.
Now, if I push my Titan card to the max, I get at most around 192.000.

I'm running +/- 3 WU's, long runs a day, which is making a big difference.
(not that it really matters, but somewhere it's a bit of a shame :-) )

Has something changed in the credit mechanism ?

Thanks to have a look !

K.

Gerard
Send message
Joined: 26 Mar 14
Posts: 101
Credit: 0
RAC: 0
Level

Scientific publications
wat
Message 42192 - Posted: 20 Nov 2015 | 9:55:28 UTC - in response to Message 42191.

The credit system has not changed. We rely on an automatic calculator of credits. Could anybody else confirm this behavior?

Vagelis Giannadakis
Send message
Joined: 5 May 13
Posts: 187
Credit: 349,254,454
RAC: 0
Level
Asp
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 42193 - Posted: 20 Nov 2015 | 10:21:58 UTC
Last modified: 20 Nov 2015 | 10:22:54 UTC

Current WU types' credit yield in credit / sec, as calculated for my GTX 970:

gerard: 6,45
sdoerr: 4,76
noelia: 5,59


I don't find this strange, it's how GPUGRID has been working since I joined (~2 years now). What I do find strange is that all WU types seem to require the same number of numeric operations, 5000000 GFLOPs.
____________

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2356
Credit: 16,377,923,309
RAC: 3,473,528
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 42194 - Posted: 20 Nov 2015 | 12:22:06 UTC - in response to Message 42193.
Last modified: 20 Nov 2015 | 12:23:00 UTC

Long runs since april this year yielded about 255.000 credits.
Now, if I push my Titan card to the max, I get at most around 192.000.

Those workunits were from a different batch, so this is normal.
Every batch has its own fixed credits, regardless its running time.
The running time depends on many factors, not just the processing speed of the GPU.
The more interaction between the GPU and the CPU the larger the runtime will be.
The credits awarded is based on how many FP operations needed for processing.
So the credits / sec (or credits / day, from which RAC is calculated) could be different between batches.

Current WU types' credit yield in credit / sec, as calculated for my GTX 980 under Windows XP x64:
GERARD_FXCXCL12_LIG_001_558_7582: 9.53 GERARD_CXCL12_DIM_HEP: 9.85 NOELIA_c36x467x: 7.36

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 42198 - Posted: 20 Nov 2015 | 21:37:15 UTC - in response to Message 42193.

What I do find strange is that all WU types seem to require the same number of numeric operations, 5000000 GFLOPs.


That's because the project doesn't bother adjusting the GFlops for every new WU type. Boinc gets there in the end, mostly.
____________
FAQ's

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

Post to thread

Message boards : Number crunching : Long runs yielding less credits ?

//