Message boards :
Number crunching :
Why the diff on BOINC CPU Benchmarks?
Message board moderation
Author | Message |
---|---|
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
I just noticed the following: 2 machines have the same CPU, yet have 20% difference on the BOINC CPU benchmarks: Computer 7158111 (other machine) Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz [Family 6 Model 60 Stepping 3] Measured floating point speed 4820.23 million ops/sec Measured integer speed 17988.95 million ops/sec Computer 7138752 (my machine) Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz [Family 6 Model 60 Stepping 3] Measured floating point speed 4005.99 million ops/sec Measured integer speed 15098.11 million ops/sec About a 20% difference!!! Any idea why? These are CPU benchmarks, remember, and so graphics cards should have no effect on them. Maybe RAM speed might be different, but these CPUs have a lot of cache, so system RAM shouldn't affect them much, I would think (depending on how the benchmarks are coded, of course). |
jason_gee Send message Joined: 24 Nov 06 Posts: 7489 Credit: 91,093,184 RAC: 0 |
Any chance there are cooling/temperature differences between the two machines ? [ also could he be overclocked ? 20% seems achievable with those ] You can monitor the clocks with CPU-Z or similar to see if the rate is dropping either due to temperatures or other power setting. Also, assuming Windows 7 (not sure where in 8.1 etc), in the OS advanced power settings there is an entry for the minimum CPU power state. For dedicated crunching, gaming etc, where the cooling and power are sufficient I set that minimum to 100%. Side note: It does seem to have an effect on feeding the GPU more promptly if the CPU is awake (logically), which is via the DPC (software interrupt) mechanims. There can be noticeable DPC latency spikes feeding the GPU of the CPU is clocked down. "Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions. |
Brent Norman Send message Joined: 1 Dec 99 Posts: 2786 Credit: 685,657,289 RAC: 835 |
The 2 computers you have on the seti list are NOT the same. So expect different results. |
Keith J. LaGue Send message Joined: 17 May 99 Posts: 59 Credit: 40,441,387 RAC: 0 |
My benchmarks dropped considerably after upgrading BOINC 7.4.36 > 7.4.42 Keifer |
Rasputin42 Send message Joined: 25 Jul 08 Posts: 412 Credit: 5,834,661 RAC: 0 |
Do you have other things running? |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
Do you have other things running? Nope, it is a dedicated cruncher. |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
The 2 computers you have on the seti list are NOT the same. So expect different results. I know that, but the difference strikes me as too large for just that. Unless the benchmark forces use of RAM, I can't see where that would make a significant difference. But then they wouldn't be good CPU benchmarks. Same CPU and likely same Intel chipsets, right? |
WezH Send message Joined: 19 Aug 99 Posts: 576 Credit: 67,033,957 RAC: 95 |
Another one has 8 processors, maybe it has HT enabled? |
HAL9000 Send message Joined: 11 Sep 99 Posts: 6534 Credit: 196,805,888 RAC: 57 |
The 2 computers you have on the seti list are NOT the same. So expect different results. The only purpose of the BOINC benchmark is to get a rough idea of the processor speed to know how much work to request. Beyond that it can be ignored. I let BOINC runs it's initial benchmark to set everything that it needs & then run with benchmarks disabled after that. SETI@home classic workunits: 93,865 CPU time: 863,447 hours Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[ |
©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.