Message boards :
Number crunching :
Benchmarking between PCs
Message board moderation
Author | Message |
---|---|
Rick Send message Joined: 23 Jun 99 Posts: 7 Credit: 810,022 RAC: 0 |
Any Ideas why I’m getting these differences? Benchmark Results – PC1: Number of CPUs: 2 1325 floating point MIPS (Whetstone) per CPU 1585 integer MIPS (Dhrystone) per CPU Benchmark Results – PC2: Number of CPUs: 2 789 floating point MIPS (Whetstone) per CPU 843 integer MIPS (Dhrystone) per CPU Both PCs using Intel P4 3,0 (socket 478) with 1GB RAM. Both have Intel Chip Set. One is AsRock and the other MSI. Why the big difference in performance? |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
It's not uncommon to have a 'bad' benchmark run once in awhile. Also, they can be highly dependant on what other software is running in the background and other factors. That being said, I have the same processor running on different MD platforms and see much benchmarks between them. So I guess the only thing you can do is to try starting Windows in a minmum configuration which is the same on both and then investigate from there to see what causing the difference (assuming that the difference has been consistently large over time). Alinator |
ohiomike Send message Joined: 14 Mar 04 Posts: 357 Credit: 650,069 RAC: 0 |
Is the slow one the MSI? MSI motherboards are a mystery to me. I have one the is running an AMD 4200+ x2 that is fine crunching, but the GUI response is horrible. I've got almost an identical machine on an ASUS MB and it is fine. (PS- both are Linux). Boinc Button Abuser In Training >My Shrubbers< |
Astro Send message Joined: 16 Apr 02 Posts: 8026 Credit: 600,015 RAC: 0 |
Rick, it seems like the cpu frequency for the lower one might be being cut in half. This can happen if overheating or if some powersaving software is kicking it (like cool'n'quiet). Try running cpuz on each and see what the operating freq is? |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
Wow, enter that post into the Bad Sentence Structure "Hall of Shame"! How could could I possibly miss that!! :-O Alinator |
OzzFan Send message Joined: 9 Apr 02 Posts: 15691 Credit: 84,761,841 RAC: 28 |
Wow, enter that post into the Bad Sentence Structure "Hall of Shame"! Hey, Alinator, don't tell anybody, but I bet you're not perfect! ;-) Shhhh. It'll be our secret. (From someone who has the occasional bad grammar/poor sentence structure post) |
Alinator Send message Joined: 19 Apr 05 Posts: 4178 Credit: 4,647,982 RAC: 0 |
:-) You caught me! I'm only 'Mary Poppins' Class. :-D Alinator |
Rick Send message Joined: 23 Jun 99 Posts: 7 Credit: 810,022 RAC: 0 |
Astro: Thanks for the tip concerning CPUZ. Ran it on both machines and discovered that the BIOS default settings on the slower machine (AsRock) were driving the DRAM at 133 instead of 200. Corrected that and matched the clock settings. Re-Ran CPUZ on it and still had the same low MIPS. Also discovered I had made a mistake in the original posting: PC1 is an ASUS with 775 socket PC2 is an AsRock with 478 socket No power saving software installed on either machine. Both XP Pro with all updates/patches. Both have i865 chip sets. Alinator: PC1 with the higher MIPS is actually running more background processes than PC2. The only difference I see right now is the 775 -vs- 478 socket. Could the benchmark software be testing the two sockets differently? Seems wierd to me! |
ohiomike Send message Joined: 14 Mar 04 Posts: 357 Credit: 650,069 RAC: 0 |
You might want to run the Cache Latency tests that come with CPUz. There could be a difference in the cache speed between the CPU's. That will make a big difference in the scores. Boinc Button Abuser In Training >My Shrubbers< |
W-K 666 Send message Joined: 18 May 99 Posts: 19136 Credit: 40,757,560 RAC: 67 |
Astro: Thanks for the tip concerning CPUZ. Ran it on both machines and discovered that the BIOS default settings on the slower machine (AsRock) were driving the DRAM at 133 instead of 200. Corrected that and matched the clock settings. Re-Ran CPUZ on it and still had the same low MIPS. Also discovered I had made a mistake in the original posting: The cpu in the 775 skt will be newer than the 478 skt cpu. Therefore it will probably have more L2 cache, always good for Seti, and the 775 pin skts connect to the northbridge and therefore to RAM with more connections allowing more data to be transfer at same clock speed. Looking at two similar units, AR = 0.3939, on your computers; resultid=545555518 This one, I assume, is done on the skt 775 computer in 31385.546875 sec resultid=548336189 And this one on the skt 478 machine in 59710.4375 sec. The results probably reflect the benchmark scores. Andy |
©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.