mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Software

Reply
 
Thread Tools
Old 2009-05-12, 11:29   #1
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

2×3×683 Posts
Default Wrong CPU speed reported in mprime benchmark

Another user sent me these benchmarks:
Quote:
running mprime under openSUSE 11.1 , kernel 2.6.27.9
Code:
AMD Athlon(tm) 64 X2 Dual Core Processor 6000+
CPU speed: 1000.55 MHz, 2 cores
CPU features: RDTSC, CMOV, Prefetch, 3DNow!, MMX, SSE, SSE2
L1 cache size: 64 KB
L2 cache size: 1 MB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 512
Prime95 64-bit version 25.9, RdtscTiming=1
Best time for 768K FFT length: 26.451 ms.
...
Note that the AMD X2 6000+ runs at 3.0GHz (and I confirmed with the user that this was the case -- Prime95 under WINE on the same machine shows 2997.64MHz). Benchmark data is also consistent with running at 3.0GHz.

I'm mostly concerned that if "faulty" benchmark data is being sent back to the server then the collection of benchmark data will get "polluted". In this case the CPU speed is off by 300% so it's fairly easy to see whether it makes sense or not, but a less drastic difference could easily slip by.
James Heinrich is offline   Reply With Quote
Old 2009-05-12, 12:05   #2
ET_
Banned
 
ET_'s Avatar
 
"Luigi"
Aug 2002
Team Italia

10010111110012 Posts
Default

Quote:
Originally Posted by James Heinrich View Post
Another user sent me these benchmarks:Note that the AMD X2 6000+ runs at 3.0GHz (and I confirmed with the user that this was the case -- Prime95 under WINE on the same machine shows 2997.64MHz). Benchmark data is also consistent with running at 3.0GHz.

I'm mostly concerned that if "faulty" benchmark data is being sent back to the server then the collection of benchmark data will get "polluted". In this case the CPU speed is off by 300% so it's fairly easy to see whether it makes sense or not, but a less drastic difference could easily slip by.
Apart from the processor speed detection, are the benchmarks correct?

Luigi
ET_ is offline   Reply With Quote
Old 2009-05-12, 12:49   #3
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

2·3·683 Posts
Default

As far as I can tell, benchmarks are completely accurate (assuming the CPU is running at 3.0GHz)
James Heinrich is offline   Reply With Quote
Old 2009-05-13, 04:37   #4
geoff
 
geoff's Avatar
 
Mar 2003
New Zealand

13·89 Posts
Default

Is frequency scaling active on this machine? One scenario that could cause an incorrect CPU speed report is where the machine is idle and running in at low frequency, then prime95 starts and records that low frequency, then the OS notices that programs are running and ramps up to high frequency, then the benchmark is run.
geoff is offline   Reply With Quote
Old 2009-05-20, 12:31   #5
James Heinrich
 
James Heinrich's Avatar
 
"James Heinrich"
May 2004
ex-Northern Ontario

100216 Posts
Default

The user (who I guess reads the forum but doesn't want to post here) said:
Quote:
Geoff was right: I disabled the "quiet and cool" option in the BIOS setting and now mprime shows 2997.54 Hz.
James Heinrich is offline   Reply With Quote
Old 2009-06-13, 12:56   #6
willmore
 
willmore's Avatar
 
Aug 2002

61 Posts
Default

I have this same issue. When starting Prime95 on a machine for the first time, I have to load the CPU to force C&Q to run the CPU at full speed. It appears that prime95 queries /proc/cpuinfo before doing much work. All that it would need to do to fix this would be to read that information after loading the CPU for a second or two.
willmore is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
mprime benchmark tests backwards? PerformanceTest Software 4 2017-03-01 14:15
NVIDIA Quadro K4000 speed results benchmark sixblueboxes GPU Computing 3 2014-07-17 00:25
Wrong CPU Speed detected w/25.8 TheOutcaste Information & Answers 7 2009-01-04 03:29
laptop reporting wrong clock speed to PrimeNet ixfd64 Hardware 1 2008-10-19 03:20
no speed gain in Prime95's benchmark when overclocking forcemaker Hardware 8 2006-04-02 15:18

All times are UTC. The time now is 01:44.


Tue Feb 7 01:44:40 UTC 2023 up 172 days, 23:13, 1 user, load averages: 0.93, 1.18, 1.36

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2023, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.

≠ ± ∓ ÷ × · − √ ‰ ⊗ ⊕ ⊖ ⊘ ⊙ ≤ ≥ ≦ ≧ ≨ ≩ ≺ ≻ ≼ ≽ ⊏ ⊐ ⊑ ⊒ ² ³ °
∠ ∟ ° ≅ ~ ‖ ⟂ ⫛
≡ ≜ ≈ ∝ ∞ ≪ ≫ ⌊⌋ ⌈⌉ ∘ ∏ ∐ ∑ ∧ ∨ ∩ ∪ ⨀ ⊕ ⊗ 𝖕 𝖖 𝖗 ⊲ ⊳
∅ ∖ ∁ ↦ ↣ ∩ ∪ ⊆ ⊂ ⊄ ⊊ ⊇ ⊃ ⊅ ⊋ ⊖ ∈ ∉ ∋ ∌ ℕ ℤ ℚ ℝ ℂ ℵ ℶ ℷ ℸ 𝓟
¬ ∨ ∧ ⊕ → ← ⇒ ⇐ ⇔ ∀ ∃ ∄ ∴ ∵ ⊤ ⊥ ⊢ ⊨ ⫤ ⊣ … ⋯ ⋮ ⋰ ⋱
∫ ∬ ∭ ∮ ∯ ∰ ∇ ∆ δ ∂ ℱ ℒ ℓ
𝛢𝛼 𝛣𝛽 𝛤𝛾 𝛥𝛿 𝛦𝜀𝜖 𝛧𝜁 𝛨𝜂 𝛩𝜃𝜗 𝛪𝜄 𝛫𝜅 𝛬𝜆 𝛭𝜇 𝛮𝜈 𝛯𝜉 𝛰𝜊 𝛱𝜋 𝛲𝜌 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔