Forum > GPU crunching

x38g reports

<< < (31/58) > >>

Jason G:

--- Quote from: perryjay on 02 Jul 2011, 03:26:49 pm ---I've only got two errors, both -12s. This one is interesting though  http://setiathome.berkeley.edu/workunit.php?wuid=770546468  Take a look at the one that did complete it. 23 triplets found??

--- End quote ---

Sure!   if ||| is one triplet, |||| is 2 triplets, and ||||| is 5 triplets.  Add a few more |'s in the same 'PulsePoT' and you get to 23 pretty quickly, which is bigger than where the nVidia code commits suicide.

I've yet to hear a good explanation of why ET might not think that's a good way to catch our attention, so do want to rewrite that to eliminate the -12s.  It's only relatively recently  that I feel my Cuda experience is getting to the point where I can consider that particular rewrite (among others), so reengineering it to match CPU is on the list, though much lower down than some other issues.

Jason

Raistmer:
Approach used in OpenCL->CUDA build shold never experience this problem. Maybe it's worth to incorporate it in "pure CUDA" build too.

Slavac:
We've got Jason sorted for a new PSU.

perryjay:
Great news Slavic. 

Get shopping Jason!!   ;D

Jason G:

--- Quote from: Raistmer on 03 Jul 2011, 01:05:41 pm ---Approach used in OpenCL->CUDA build shold never experience this problem. Maybe it's worth to incorporate it in "pure CUDA" build too.
--- End quote ---

Yeah, when I rewrite that I'll take yours into consideration & apply the 'max bandwidth' approach that's been working well for me so far as well.  As mentioned I do want that issue gone, and as errors go it's becoming more common as hardware gets faster & the work noiser.

 
--- Quote from: Slavac on 03 Jul 2011, 06:34:51 pm ---We've got Jason sorted for a new PSU.

--- End quote ---

Wow looks like I end up with enough to get the 'big one'.   :)  I'll let everyone know the good news on main & issue thanks as well. 

Navigation

[0] Message Index

[#] Next page

[*] Previous page

Go to full version