Message boards : Number crunching : Linux vs Windows point awards
Author | Message |
---|---|
arcturus Send message Joined: 22 Sep 05 Posts: 16 Credit: 525,440 RAC: 0 |
Same computer, similar work unit. Boinc linux client v5.2.5. 2,240.67 seconds 7.89 points - Windows 2,232.89 seconds 4.28 points - Linux In case anybody cares. |
The Pirate Send message Joined: 22 Sep 05 Posts: 20 Credit: 7,090,933 RAC: 0 |
|
Daddygeek Send message Joined: 17 Sep 05 Posts: 12 Credit: 3,993,423 RAC: 0 |
It might be beaten to death, but is there a fix, a remedy, something we can do? Server 2003 is not bad, but I want my Linux back. |
Tern Send message Joined: 25 Oct 05 Posts: 576 Credit: 4,695,177 RAC: 21 |
It might be beaten to death, but is there a fix, a remedy, something we can do? About the only thing to do is install an optimized BOINC client that will improve your benchmarks. I don't have Linux, so don't have a link handy, but I'd look on the SETI boards for "optimized boinc linux" - you should find something. Or someone may have a link they'll post here. The problem right now is that with the new BOINC versions just out, there's usually a lag before someone optimizes them. |
Andrew Send message Joined: 19 Sep 05 Posts: 162 Credit: 105,512 RAC: 0 |
You can find an optimized linux client here: http://naparst.name/boinc.htm Or search the seti forum :) Mr. Kim et el, have said that they will be releasing the Rosetta client source soon so that people can help optimize/fix the client for different platforms. |
Charles Dennett Send message Joined: 27 Sep 05 Posts: 102 Credit: 2,074,126 RAC: 232 |
Or, you could download the source and compile it yourself. That's what I do. -Charlie |
Andrew Send message Joined: 19 Sep 05 Posts: 162 Credit: 105,512 RAC: 0 |
Get the source here: http://boinc.berkeley.edu/source_code.php This page will help compile it: http://boinc.berkeley.edu/compile.php |
Desti Send message Joined: 16 Sep 05 Posts: 50 Credit: 3,018 RAC: 0 |
Yea, this subject has pretty much been beaten to death. That's true, but normaly the validation process gives the Linuxers mostly more credits than they claimed for. Rosetta has no validation against Windows(tm) crunchers, so there is no credit mixing. LUE |
arcturus Send message Joined: 22 Sep 05 Posts: 16 Credit: 525,440 RAC: 0 |
Get the source here: http://boinc.berkeley.edu/source_code.php How's this going to help without Rosetta client source code? |
Divide Overflow Send message Joined: 17 Sep 05 Posts: 82 Credit: 921,382 RAC: 0 |
You don't need the Rosetta application source code to fix the claimed credit difference issue. Credit is a BOINC core client matter. The links provided will help you download and compile a linux version of BOINC that should give you better BOINC benchmarks. That should produce Linux credit scores that are closer to Windows ones (for the same hardware). |
Andrew Send message Joined: 19 Sep 05 Posts: 162 Credit: 105,512 RAC: 0 |
If you have the Rosetta client source code, what would you do with it? |
arcturus Send message Joined: 22 Sep 05 Posts: 16 Credit: 525,440 RAC: 0 |
If you have the Rosetta client source code, what would you do with it? Doh ... optimize it maybe? |
Andrew Send message Joined: 19 Sep 05 Posts: 162 Credit: 105,512 RAC: 0 |
Doh ... optimize it maybe? Well... if by optimize you mean "complete a WU faster"... then that will not work. If the linux client runs faster it will claim even less credit then it is claiming now. The amount of credit claimed is based on how long the WU took. That is why there are optimized boinc clients. The optimized boinc client increases the claimed credit from a optimized project by optimizing/increasing the benchmark results. |
Biggles Send message Joined: 22 Sep 05 Posts: 49 Credit: 102,114 RAC: 0 |
Doh ... optimize it maybe? My understanding is that with BOINC, credit = CPU time * benchmark score. Except this is averaged against all those who complete a WU, with the greatest and lowest claims discarded. For simplicity's sake, let us say that 5 users complete the same WU. Let us also say that we are running an optimised client, which took less time to complete meaning we claimed less credit. We'll claim 45 for instance. Now everyone else took longer to run the WU than we did, even if they have identical computers, because they are running standard clients. So they claim 50, because they took longer than us. And to make the example nice and easy, we'll have one guy who took longer still and claimed 55. Therefore, BOINC would discard the greatest and lowest claimed credits, 45 and 55 in our example, and award the average of the remaining claimed credits, this being 50. So we got 50 credits in less time than everyone else, because we used an optimised client. Does that make sense? Or am I just wrong? |
Housing and Food Services Send message Joined: 1 Jul 05 Posts: 85 Credit: 155,098,531 RAC: 0 |
Rosetta doesn't repeat work units since they are able to tell if a result is forged due to the nature of their code. This causes whatever credit a client claims to be granted. . so it's a 'problem' in the Boinc application, not Rosetta itself. -E |
Biggles Send message Joined: 22 Sep 05 Posts: 49 Credit: 102,114 RAC: 0 |
OK well in that case ignore my previous post. However, with regard to redundancy, is there any performed? I wouldn't feel particularly comfortable with there not being any since all it would take is an unstable computer to give bad results leading to bad research. At least if there is some redundancy you can pick up on that. |
arcturus Send message Joined: 22 Sep 05 Posts: 16 Credit: 525,440 RAC: 0 |
Well... if by optimize you mean "complete a WU faster"... then that will not work. If the linux client runs faster it will claim even less credit then it is claiming now. The amount of credit claimed is based on how long the WU took. Underclock your pc and benchmark it to increase credit claimed, then crank it up. The increased number of wu's thanks to optimized Rosetta code produces greater overall point production ... unless there's a proportionate decrease in the benchmark score. |
Tern Send message Joined: 25 Oct 05 Posts: 576 Credit: 4,695,177 RAC: 21 |
Underclock your pc and benchmark it to increase credit claimed, then crank it up. Rube Goldberg was famous for things like that... doing something the INCREDIBLY complicated way, rather than the blatently simple way... not even mentioning the obvious "run an optimized BOINC client". The problem here is we're talking about Rosetta. No optimized applications, thus no reason to mess with the benchmarks, unless they are way low. |
Andrew Send message Joined: 19 Sep 05 Posts: 162 Credit: 105,512 RAC: 0 |
Underclock your pc and benchmark it to increase credit claimed, then crank it up. This won't work either because the boinc client benchmarks the computer periodically throughout the day. So you'll have to sit watching your computer 24/7 to catch it doing the benchmarks. ... so it's a 'problem' in the Boinc application, not Rosetta itself. I find this really interesting... Out of curiosity if someone has a dual boot win/linux on the machine, maybe they could posts some benchmarks numbers to see if the boinc client is reporting similar Whetstone and Dhrystone numbers. If it's a boinc client problem then the linux benchmark will be significantly lower than the window's ones. |
arcturus Send message Joined: 22 Sep 05 Posts: 16 Credit: 525,440 RAC: 0 |
Underclock your pc and benchmark it to increase credit claimed, then crank it up. Actually it's been 24 hours so far since the last benchmark. |
Message boards :
Number crunching :
Linux vs Windows point awards
©2024 University of Washington
https://www.bakerlab.org