Message boards :
Number crunching :
Whetstone on Dual boot Linux - Windows differ a factor
Message board moderation
Author | Message |
---|---|
Huibert Kreb Send message Joined: 3 Apr 99 Posts: 7 Credit: 234,955 RAC: 0 |
How can one machine give different Whetstone numbers when running under Linux versus Windows? Should these numbers not be the same PER DEFINITION? And the WU processing time is equal on both instances... Netvista onder windows: (Windows XP Home) Number of CPUs: 1 1350 double precision MIPS (Whetstone) per CPU 2403 integer MIPS (Dhrystone) per CPU Netvista onder Linux: (RHEL 3.0) Number of CPUs: 1 575 double precision MIPS (Whetstone) per CPU 1193 integer MIPS (Dhrystone) per CPU Any suggestions? Huibert |
G.A.Waldenmeier@home_Project Send message Joined: 3 Apr 99 Posts: 63 Credit: 0 RAC: 0 |
maybe people build boinc on M$ 7.0 C++ Compiler [/url] [/url] Phantasie ist wichtiger als Wissen,denn Wissen ist begrenzt.Albert Einstein |
guidoprost Send message Joined: 3 Apr 99 Posts: 7 Credit: 19,732 RAC: 0 |
I think the problem is that the Linux client is compiled without any optimizations turned on at all. I experimented a little bit a while ago and posted a little howto here: http://setiweb.ssl.berkeley.edu/forum_thread.php?id=1350 My benchmarks nearly doubled, it is still not the same as Windows on my dual boot machine, but at least it's an improvement. Just make sure you find out about the right compiler flags for your computer first. BTW, I have received credit since I switched to my self-compiled version, so that's no problem as long as you don't use -ffast-math or something else that messes with precision. Another hint: don't use the newest sources (version 4+), they won't work yet. I think I'm using 07-11 boinc and 07-12 seti at the moment. |
©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.