Message boards :
Number crunching :
Benchmarking!
Message board moderation
Author | Message |
---|---|
The Gas Giant Send message Joined: 22 Nov 01 Posts: 1904 Credit: 2,646,654 RAC: 0 |
Does the new development stream of BOINC V4.6x etc benchmark windows boxes properly or are we still going to be claiming significantly different credit for the same work crunched? Paul (S@H1 8888) And proud of it! |
Steve Dundes Send message Joined: 6 Sep 04 Posts: 43 Credit: 159,057 RAC: 0 |
I saw somewhere in a different thread that there was a bug in the benchmarking for BOINC 4.1x and it was fixed in 4.5x and the fix was carried through to 4.66. Basically what I have found is you will receive less credit with the 4.6x version. I had some free time and put together a spreadsheet for the benchmarks and compairing versions 4.19, 4.54, 4.62 and 4.66 for the six different cpu's that I run on various projects. As expected the 4.5x and up all scored the same but the difference was in 4.19. For example I have a server that has a P3 550Mhz processor and it takes about 12 hours to run a Seti unit. Under 4.19 I would claim a credit of 42.68 but under versions 4.5x and up I would only claim a credit of 33.63. Still not sure if the benchmark for windows in 4.6x is proper but it is less than what you are using now. |
Professor Desty Nova Send message Joined: 17 May 99 Posts: 59 Credit: 579,918 RAC: 0 |
From BOINC CVS checkin notes in November 2004: David 2 Nov 2004 - split Dhrystone source into 2 files (from Peter Smithson). This supposedly prevents compiler optimizations on Win that give inflated results From BOINC_dev mailing List (Peter Smithson): The original dhrystone source for revison 2.0 was split intentionally to avoid over optimisation (see comments in original dhrystone test source code). The BOINC uses merged source which causes the benchmark to run unusually quickly when compiled with MSVC 7.1 due to better optimisation possible. Presumably the optimiser is taking out most of the test and not giving a result consistent with a real application. MSVC 6.0 did not do this. Link SETI@home classic workunits: 1,985 CPU time: 24,567 hours Professor Desty Nova Researching Karma the Hard Way |
The Gas Giant Send message Joined: 22 Nov 01 Posts: 1904 Credit: 2,646,654 RAC: 0 |
Cool...thanks guys. I wonder if CPDN will drop it's trickle credit back to 75 from 95 once we all go the current dev version? |
©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.