mersenneforum.org  

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

Reply
 
Thread Tools
Old 2003-09-01, 21:48   #23
delta_t
 
delta_t's Avatar
 
Nov 2002
Anchorage, AK

3×7×17 Posts
Default

Another benchmark for a Pentium-M 1.4 GHz (Centrino) from a Compaq Presario X1000 laptop.



Intel(R) Pentium(R) M processor 1400MHz
CPU speed: 598.04 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 128
Prime95 version 23.6, RdtscTiming=1
Best time for 384K FFT length: 42.841 ms.
Best time for 448K FFT length: 51.516 ms.
Best time for 512K FFT length: 57.944 ms.
Best time for 640K FFT length: 74.285 ms.
Best time for 768K FFT length: 90.659 ms.
Best time for 896K FFT length: 107.811 ms.
Best time for 1024K FFT length: 120.382 ms.
Best time for 1280K FFT length: 153.983 ms.
Best time for 1536K FFT length: 186.577 ms.
Best time for 1792K FFT length: 221.972 ms.
Best time for 2048K FFT length: 253.494 ms.
delta_t is offline   Reply With Quote
Old 2003-09-02, 00:30   #24
tom11784
 
tom11784's Avatar
 
Aug 2003
Upstate NY, USA

2·163 Posts
Default

I had a similar problem with my laptop at first:

My default power scheme had the laptop processor running at "Medium" which produced numbers like:

[code:1]Intel(R) Pentium(R) 4 Mobile CPU 1.80GHz
CPU speed: 1196.13 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 8 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 version 22.12, RdtscTiming=1
Best time for 256K FFT length: 21.5 ms.
Best time for 320K FFT length: 27.9 ms.
Best time for 384K FFT length: 34.4 ms.
Best time for 448K FFT length: 41.8 ms.
Best time for 512K FFT length: 46.8 ms.
Best time for 640K FFT length: 61.2 ms.
Best time for 768K FFT length: 74.4 ms.
Best time for 896K FFT length: 93.7 ms.
Best time for 1024K FFT length: 96.4 ms.
Best time for 1280K FFT length: 142.9 ms.
Best time for 1536K FFT length: 177.6 ms.
Best time for 1792K FFT length: 219.2 ms.[/code:1]


Then I entered my "Battery MaxiMiser Wizard" and created my own power setting where my CPU runs at Maximum, producing numbers such as:

[code:1]Intel(R) Pentium(R) 4 Mobile CPU 1.80GHz
CPU speed: 1794.07 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 8 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 64
Prime95 version 22.12, RdtscTiming=1
Best time for 256K FFT length: 14.4 ms.
Best time for 320K FFT length: 19.6 ms.
Best time for 384K FFT length: 23.4 ms.
Best time for 448K FFT length: 27.7 ms.
Best time for 512K FFT length: 31.9 ms.
Best time for 640K FFT length: 40.3 ms.
Best time for 768K FFT length: 50.7 ms.
Best time for 896K FFT length: 64.9 ms.
Best time for 1024K FFT length: 67.0 ms.
Best time for 1280K FFT length: 96.9 ms.
Best time for 1536K FFT length: 121.8 ms.
Best time for 1792K FFT length: 152.1 ms.[/code:1]

Hope this helps. Also, on my IBM T30, you enter the Wizard by double-clicking on the Green bar that says what percentage of power your battery has. Lemme know if this works for anyone.

-Tom
tom11784 is offline   Reply With Quote
Old 2003-09-02, 10:14   #25
czolus
 
Sep 2003

32 Posts
Default

The low Banias timings are a direct result of the chip having SSE2 latencies about 1.5x->2x slower than an equivalent P4. Likewise, the P4 added significant latencies to many regular fpu instructions to make space for better SSE2 latencies. Banias is designed more for office-style applications, and less for heavy scientific (and gaming) work. The actual slowdown is lower clock frequency * higher latencies. The better IPC really only effects the alu's and fpu.[/i]

Another thing to note is that the heavy use of prefetch instructions basically negates the advantage of having a larger cache on the Banias. I won't get into the details, but that explains why the pitiful Celeron4 gets decent scores in Prime95.
czolus is offline   Reply With Quote
Old 2003-09-02, 23:44   #26
delta_t
 
delta_t's Avatar
 
Nov 2002
Anchorage, AK

3·7·17 Posts
Default

Hello tom11784,

I noticed your laptop processor may not be the Centrino processor we are talking about (take a look at your L2 cache). I believe the new Centrino (Pentium-M's) have an L2 cache of 1024K.

I had all power savings software disabled when I ran the benchmarks. Hopefully at some point this will be fixed.
delta_t is offline   Reply With Quote
Old 2003-09-03, 05:25   #27
tom11784
 
tom11784's Avatar
 
Aug 2003
Upstate NY, USA

1010001102 Posts
Default

I know it isn't a Centrino, but I figured if I missed it, even though I'm no genius or anything, that perhaps someone else missed it when they received their new laptop.
tom11784 is offline   Reply With Quote
Old 2003-09-21, 07:00   #28
delta_t
 
delta_t's Avatar
 
Nov 2002
Anchorage, AK

3×7×17 Posts
Default

Here is a 1GHz Centrino benchmark with 23.7


Intel(R) Pentium(R) M processor 1000MHz
CPU speed: 1000.01 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE, SSE2
L1 cache size: 32 KB
L2 cache size: 1024 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
TLBS: 128
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 58.234 ms.
Best time for 448K FFT length: 71.127 ms.
Best time for 512K FFT length: 79.206 ms.
Best time for 640K FFT length: 102.371 ms.
Best time for 768K FFT length: 125.058 ms.
Best time for 896K FFT length: 144.915 ms.
Best time for 1024K FFT length: 165.598 ms.
Best time for 1280K FFT length: 212.396 ms.
Best time for 1536K FFT length: 253.581 ms.
Best time for 1792K FFT length: 324.428 ms.
Best time for 2048K FFT length: 344.232 ms.
delta_t is offline   Reply With Quote
Old 2003-09-21, 07:04   #29
delta_t
 
delta_t's Avatar
 
Nov 2002
Anchorage, AK

16516 Posts
Default

I know George is probably busy working on the Opteron/Athlon 64 tweaks of Prime95, but just want to see what kind of computer access George would need to get the performance of the Centrinos up to the Pentium 4. I suppose I may be able to provide access to a Centrino machine, depending on what kind of access George would need.

Thanks,
delta_t is offline   Reply With Quote
Old 2003-11-03, 23:37   #30
magicfan241
 
Nov 2003

7 Posts
Default

If anyone wants, I can run any pre-complied source (or if you can teach me how to complie the stuff (and tell me where to get free stuff to use)) on my Centrino laptop.

I can't give a remote desktop, because other people use this computer heavily, and they might notice.

Whatever would help you the best is fine with me, but I am a coding n00b (except in JAVA, but that does not apply here)

you can reach me at: magicfan241 at netscape dot net

or on the Ars forums, in the DCA.

magicfan241
magicfan241 is offline   Reply With Quote
Old 2003-11-04, 08:29   #31
Dresdenboy
 
Dresdenboy's Avatar
 
Apr 2003
Berlin, Germany

36110 Posts
Default

Quote:
Originally posted by magicfan241
If anyone wants, I can run any pre-complied source (or if you can teach me how to complie the stuff (and tell me where to get free stuff to use)) on my Centrino laptop.
I mailed you a small instruction timing benchmark. I think it will be interesting to see the measured numbers.

And right before sending this post I remembered another test which you could run. It measures the speed of reading from cache (mail is sent).
Dresdenboy is offline   Reply With Quote
Old 2003-11-04, 14:46   #32
magicfan241
 
Nov 2003

7 Posts
Default

here are the results to one program you sent me.

CPU features:a7e9f9bf
Unit Latency Throghput
ALU 1.01 0.52
IMUL 4.00 1.00
ISHIFT 1.00 1.00
ISHIFT(NetBurst Opt.) 1.01 1.00
x87 ADD 3.00 1.05
x87 MUL 5.00 2.02
MMX ADD 1.01 0.51
MMX MUL 3.00 1.05
SSE Scalar SP ADD 3.00 1.05
SSE Packed SP ADD 3.01 2.05
SSE Scalar SP MUL 4.00 1.01
SSE Packed SP MUL 4.00 2.01
SSE2 Scalar DP ADD 3.00 1.04
SSE2 Packed DP ADD 3.01 2.05
SSE2 Scalar DP MUL 5.00 2.00
SSE2 Packed DP MUL 5.00 4.01
SSE2 Packed INT ADD 1.18 2.01
SSE2 Packed INT MUL 3.01 2.05

I clicked run, and waited for my activity light to stop blinking, and found this file, creativly called results.txt

Hope this helps

magicfan241
magicfan241 is offline   Reply With Quote
Old 2003-11-04, 14:48   #33
magicfan241
 
Nov 2003

710 Posts
Default

here are the resluts to the second program you sent me:

Running MMX/SSE/SSE2 reading speed tests... a7e9f9bf

Time for reading 128kB from cache using MOVQ: 28676 cycles (4.57 Bytes/cycle)
Time for reading 8kB from cache using MOVQ: 1076 cycles (7.61 Bytes/cycle)
Time for reading 128kB from cache using MOVAPD: 28676 cycles (4.57 Bytes/cycle)
Time for reading 8kB from cache using MOVAPD: 1076 cycles (7.61 Bytes/cycle)
Time for reading 128kB from cache using MOVDQA: 28676 cycles (4.57 Bytes/cycle)
Time for reading 8kB from cache using MOVDQA: 1075 cycles (7.62 Bytes/cycle)
Time for reading 128kB from cache using MOVAPS: 28672 cycles (4.57 Bytes/cycle)
Time for reading 8kB from cache using MOVAPS: 1079 cycles (7.59 Bytes/cycle)


magicfan241
magicfan241 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
29.2 benchmark help Prime95 Software 69 2017-05-23 23:49
Benchmark Estimate Primeinator Information & Answers 8 2009-06-11 23:39
Does anyone have i7 920? for Benchmark? cipher Twin Prime Search 2 2009-04-14 20:16
Not happy on Centrino delta_t NFSNET Discussion 7 2004-01-09 16:03
Centrino has problems with Prime95... magicfan241 Software 1 2003-11-03 20:46

All times are UTC. The time now is 12:58.


Mon May 16 12:58:24 UTC 2022 up 32 days, 10:59, 1 user, load averages: 1.59, 1.43, 1.47

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2022, 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 𝛴𝜎 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔