Invalid workunit (5 invalids on GPU valid on 2 CPUs)

Message boards : Number crunching : Invalid workunit (5 invalids on GPU valid on 2 CPUs)
Message board moderation

Author	Message
BetelgeuseFive Volunteer tester Send message Joined: 6 Jul 99 Posts: 158 Credit: 17,117,787 RAC: 19	Message 1746100 - Posted: 1 Dec 2015, 13:46:22 UTC Hi folks, Long time since I had an invalid. This one seems interesting as it was marked invalid on 5 different GPUs. It validated on 2 CPUs. http://setiathome.berkeley.edu/workunit.php?wuid=1954114167 Any clues anyone ? Tom ID: 1746100 ·
Richard Haselgrove Volunteer tester Send message Joined: 4 Jul 99 Posts: 14650 Credit: 200,643,578 RAC: 874	Message 1746102 - Posted: 1 Dec 2015, 14:06:19 UTC - in response to Message 1746100.   That's one of the WUs from the faulty splitter run on 4 November. The CUDA apps are explicit about what they were doing: "Legacy setiathome_enhanced V6 mode". The intel_gpu seems to have freaked out, and the Mac ATI crashed (SIGSEGV: segmentation violation). The surprise is perhaps that there were enough stock CPU attempts to validate. ID: 1746102 ·
Jeff Buck Volunteer tester Send message Joined: 11 Feb 00 Posts: 1441 Credit: 148,764,870 RAC: 0	Message 1746118 - Posted: 1 Dec 2015, 15:51:05 UTC - in response to Message 1746102.   Last modified: 1 Dec 2015, 15:56:46 UTC I've seen half a dozen or so like that from the November 4 batch that passed through here. They're all -9 overflows where the validator doesn't raise the instant Invalid due to the Autocorr being MIA. EDIT: I'm thinking that all the ones I saw were VLARs, also, so they didn't require that many resends to hit a couple of stock CPU hosts. ID: 1746118 ·
BetelgeuseFive Volunteer tester Send message Joined: 6 Jul 99 Posts: 158 Credit: 17,117,787 RAC: 19	Message 1746119 - Posted: 1 Dec 2015, 15:51:58 UTC - in response to Message 1746102.   That's one of the WUs from the faulty splitter run on 4 November. The CUDA apps are explicit about what they were doing: "Legacy setiathome_enhanced V6 mode". The intel_gpu seems to have freaked out, and the Mac ATI crashed (SIGSEGV: segmentation violation). The surprise is perhaps that there were enough stock CPU attempts to validate. Thanks for the explanation, Tom ID: 1746119 ·

 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.