This is just to let you know that we are working on it. It definitely is more difficult than I first thought.
It might help if someone of you having this problem could send us (or post here, stripped of any personal info you like) the init_data.xml files from the two slots of the tasks that end up running on the same device.
BM
I'll set up a run and send you the files (email OK?) when this GPUGrid task vacates its GPU - ~2.5 hours.
I didn't want to spam the boards with my stats - just milestone theads - but apparently signatures are no longer optional. Follow the link if you're interested.
This is just to let you know that we are working on it. It definitely is more difficult than I first thought.
It might help if someone of you having this problem could send us (or post here, stripped of any personal info you like) the init_data.xml files from the two slots of the tasks that end up running on the same device.
BM
I'll set up a run and send you the files (email OK?) when this GPUGrid task vacates its GPU - ~2.5 hours.
Email is ok. But it will probably be enough to post the tags of the two files here ( and whatever I might have missed).
This is just to let you know that we are working on it. It definitely is more difficult than I first thought.
It might help if someone of you having this problem could send us (or post here, stripped of any personal info you like) the init_data.xml files from the two slots of the tasks that end up running on the same device.
BM
I'll set up a run and send you the files (email OK?) when this GPUGrid task vacates its GPU - ~2.5 hours.
Email is ok. But it will probably be enough to post the tags of the two files here ( and whatever I might have missed).
BM
Well, those tags look as I would have expected:
NVIDIA
0
0
NVIDIA
1
1
I was a little worried about a couple of 'device_id 0' lower down:
7
0
0
8
0
0
I'll zip up the files and email them so you have everything available in the same place.
I didn't want to spam the boards with my stats - just milestone theads - but apparently signatures are no longer optional. Follow the link if you're interested.
Email is ok. But it will probably be enough to post the tags of the two files here ( and whatever I might have missed).
BM
Same here:
NVIDIA
0
0
and
NVIDIA
1
1
Is it possible to add to the WU output something like:
[INFO ] Using CUDA device #0 "GeForce GTX 660 Ti" (0 CUDA cores / 0.00 GFLOPS)
as it was done for CUDA?
This is not a problem or bug but there is no other thread possibility then news.
The FGRP (gamma ray) via ATI are finished in around 1350 seconds on a old HD5870 card.
On cpu via Einstein@home (also gamma ray, I thus I think the same)it takes my pc around 8200 seconds to finish. So a good improvement.
This is not a problem or bug but there is no other thread possibility then news.
The FGRP (gamma ray) via ATI are finished in around 1350 seconds on a old HD5870 card.
On cpu via Einstein@home (also gamma ray, I thus I think the same)it takes my pc around 8200 seconds to finish. So a good improvement.
Thanks for the feedback. Yes, the good old HD 5850 and 5870 are quite fast, if only they would consume less power :-[ .
I'm quite satisfied with the performance for the moment, but there still is some work to do under the hood (fixing a bug of mine in GPU detection, reducing verbosity of log, error handling needs to be more robust ....). But I think I can put out a new fixed version tomorrow latest.
Thanks; seems a shame, because it's so close. I'll have to have a look and see if I can find out where the missing 192kb is! (it's not in use by the OS, being headless systems).
RE: This is just to let you
I'll set up a run and send you the files (email OK?) when this GPUGrid task vacates its GPU - ~2.5 hours.
I didn't want to spam the boards with my stats - just milestone theads - but apparently signatures are no longer optional. Follow the link if you're interested.
http://www.boincsynergy.com/images/stats/comb-3475.jpg
RE: RE: This is just to
Email is ok. But it will probably be enough to post the tags of the two files here ( and whatever I might have missed).
BM
BM
RE: RE: RE: This is
Well, those tags look as I would have expected:
I was a little worried about a couple of 'device_id 0' lower down:
I'll zip up the files and email them so you have everything available in the same place.
I didn't want to spam the boards with my stats - just milestone theads - but apparently signatures are no longer optional. Follow the link if you're interested.
http://www.boincsynergy.com/images/stats/comb-3475.jpg
RE: Email is ok. But it
Same here:
NVIDIA
0
0
and
NVIDIA
1
1
RE: RE: Email is ok. But
Is it possible to add to the WU output something like:
[INFO ] Using CUDA device #0 "GeForce GTX 660 Ti" (0 CUDA cores / 0.00 GFLOPS)
as it was done for CUDA?
This is not a problem or bug
This is not a problem or bug but there is no other thread possibility then news.
The FGRP (gamma ray) via ATI are finished in around 1350 seconds on a old HD5870 card.
On cpu via Einstein@home (also gamma ray, I thus I think the same)it takes my pc around 8200 seconds to finish. So a good improvement.
Greetings from,
TJ.
RE: This is not a problem
Thanks for the feedback. Yes, the good old HD 5850 and 5870 are quite fast, if only they would consume less power :-[ .
I'm quite satisfied with the performance for the moment, but there still is some work to do under the hood (fixing a bug of mine in GPU detection, reducing verbosity of log, error handling needs to be more robust ....). But I think I can put out a new fixed version tomorrow latest.
Cheers
HB
Will the new version work on
Will the new version work on cards with 512Mb, do you know? (ref. Message 112402).
RE: Will the new version
Memory requirements will be unchanged, sorry.
HB
Thanks; seems a shame,
Thanks; seems a shame, because it's so close. I'll have to have a look and see if I can find out where the missing 192kb is! (it's not in use by the OS, being headless systems).