Optimized Client Versions or Source code?

Questions and Answers : Wish list : Optimized Client Versions or Source code?

To post messages, you must log in.

AuthorMessage
D J Blumer

Send message
Joined: 6 Nov 05
Posts: 3
Credit: 52,890,723
RAC: 31,022
Message 2549 - Posted: 7 Nov 2005, 5:15:00 UTC

Have the Rosetta clients been extensively optimized for each type of processor? I am running an optimized version of SETI@Home for the Mac G5 (hand optimized Altivec vectorized code, G5 compiler optimizations, Shark, etc.) which is about 2X-3X faster than the conventional OSX client (~3000 sec vs 9000 sec/SETI WU). Would Rosetta benefit much from the hardware vector processor and/or other optimizations, or has this already been exhausted?
ID: 2549 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
J D K
Avatar

Send message
Joined: 23 Sep 05
Posts: 168
Credit: 101,266
RAC: 0
Message 2730 - Posted: 9 Nov 2005, 16:10:53 UTC

It has been discussed and the source may be released at a later date...
BOINC Wiki

ID: 2730 · Rating: 1 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dgnuff
Avatar

Send message
Joined: 1 Nov 05
Posts: 350
Credit: 24,773,605
RAC: 0
Message 3250 - Posted: 15 Nov 2005, 4:24:38 UTC - in response to Message 2730.  

It has been discussed and the source may be released at a later date...


Also, at this stage, I get the read that David Baker and the Rosetta crew are still working on the exact details of the algorithm. At least that's how his "Breakthrough" thread seems to me. Optimizing at this point is probably not high on their radar.

First make it work, THEN make it work fast. -- anonymous

The worst mistake you can make [in programming] is premature optimization -- Dan Simkins.
ID: 3250 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Questions and Answers : Wish list : Optimized Client Versions or Source code?



©2024 University of Washington
https://www.bakerlab.org