Suggestion - move benchmark code into science clients

Message boards : Number crunching : Suggestion - move benchmark code into science clients
Message board moderation

To post messages, you must log in.

AuthorMessage
Don Erway
Volunteer tester

Send message
Joined: 18 May 99
Posts: 305
Credit: 471,946
RAC: 0
United States
Message 140011 - Posted: 21 Jul 2005, 5:50:55 UTC

It seems to me, that it would simplify life, if you moved the open source code from the BOINC CC, to the individual science clients.

That way, any compilation optimizations would automatically be picked up, and could in fact, be different, for different projects, and we would never have to mess with optimized BOINCs.

Don

ID: 140011 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19062
Credit: 40,757,560
RAC: 67
United Kingdom
Message 140039 - Posted: 21 Jul 2005, 7:25:59 UTC

Hi Don,

Already suggested, but not going to happen, but they are working on it in ver.4.7x. Which may become ver. 5.x.

Andy

I also joined 18 May 1999, but lost my password, when I moved and had to change ISP.
ID: 140039 · Report as offensive
Profile Paul D. Buck
Volunteer tester

Send message
Joined: 19 Jul 00
Posts: 3898
Credit: 1,158,042
RAC: 0
United States
Message 140169 - Posted: 21 Jul 2005, 13:19:33 UTC

An optimized BOINC Daemon/Manager will not buy you much as very little time is spent in those processes. The bulk of the time is spend in the Science Application. So, that is where optimizing would, and does, pay off.
ID: 140169 · Report as offensive
Don Erway
Volunteer tester

Send message
Joined: 18 May 99
Posts: 305
Credit: 471,946
RAC: 0
United States
Message 140211 - Posted: 21 Jul 2005, 14:38:32 UTC - in response to Message 140169.  

An optimized BOINC Daemon/Manager will not buy you much as very little time is spent in those processes. The bulk of the time is spend in the Science Application. So, that is where optimizing would, and does, pay off.


Exactly my point!

But, the benchmarks are built into the BOINC daemon, and thus, if you want benchmarks that match reality, folks are suggesting that you should run an optimized BOINC, if you run an optimized client.


ID: 140211 · Report as offensive
Don Erway
Volunteer tester

Send message
Joined: 18 May 99
Posts: 305
Credit: 471,946
RAC: 0
United States
Message 140264 - Posted: 21 Jul 2005, 16:30:27 UTC - in response to Message 140039.  


I also joined 18 May 1999, but lost my password, when I moved and had to change ISP.


I got very lucky there - found my original password on a CD, of old emacs email archives!

Don



ID: 140264 · Report as offensive

Message boards : Number crunching : Suggestion - move benchmark code into science clients


 
©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.