mersenneforum.org  

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

Reply
 
Thread Tools
Old 2003-11-05, 18:02   #1
optim
 
optim's Avatar
 
Nov 2003
European Union

23·13 Posts
Cool Benchmark: AMD T-bred B and Barton

There are three AMD Athlon XP 2600+ variants:

1. T-bred B: 2.13 GHz, 256K L2 cache, FSB266
2. T-Bred B: 2083 MHz, 256K L2 cache, FSB333
3. Barton: 1917 MHz, 512K L2 cache, FSB333

Here you can see the difference between [2] and [3]:

CPU type: AMD Athlon XP 2600+ (Barton core, FSB333)
CPU Speed: 1917 MHz
L2 cache size: 512 KB
L2 cache speed: Full
Memory Type / Bus Speed: PC2700 DDR333
OS: Windows XP Pro SP1
Notes: N/A
AMD Athlon(tm) XP 2600+
CPU speed: 1913.31 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 31.301 ms.
Best time for 448K FFT length: 36.159 ms.
Best time for 512K FFT length: 38.922 ms.
Best time for 640K FFT length: 53.540 ms.
Best time for 768K FFT length: 62.554 ms.
Best time for 896K FFT length: 76.862 ms.
Best time for 1024K FFT length: 85.073 ms.
Best time for 1280K FFT length: 111.915 ms.
Best time for 1536K FFT length: 132.184 ms.
Best time for 1792K FFT length: 161.942 ms.
Best time for 2048K FFT length: 178.698 ms.



CPU type: AMD Athlon XP 2600+ 2083MHz 256K L2 Cache FSB333
CPU Speed: 2083
L2 cache size: 256
L2 cache speed: Full
Memory Type / Bus Speed: PC2700 DDR333
OS: Windows Server 2003
Notes: client version 23.7
AMD Athlon(tm) XP 2600+
CPU speed: 2088.01 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 256 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 30.321 ms.
Best time for 448K FFT length: 34.637 ms.
Best time for 512K FFT length: 37.770 ms.
Best time for 640K FFT length: 49.844 ms.
Best time for 768K FFT length: 59.188 ms.
Best time for 896K FFT length: 70.837 ms.
Best time for 1024K FFT length: 79.076 ms.
Best time for 1280K FFT length: 105.576 ms.
Best time for 1536K FFT length: 124.971 ms.
Best time for 1792K FFT length: 155.891 ms.
Best time for 2048K FFT length: 177.904 ms.

Last fiddled with by optim on 2003-11-05 at 18:03
optim is offline   Reply With Quote
Old 2003-11-06, 08:32   #2
optim
 
optim's Avatar
 
Nov 2003
European Union

23·13 Posts
Default

This benchmark is for my AMD Athlon XP 2600+ 1.91GHz Barton 512K Cache overclocked to 2800+ speeds with the same FSB speed (FSB333).

AMD Athlon(tm) XP 2800+
CPU speed: 2079.49 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 30.300 ms.
Best time for 448K FFT length: 35.154 ms.
Best time for 512K FFT length: 37.531 ms.
Best time for 640K FFT length: 49.894 ms.
Best time for 768K FFT length: 59.205 ms.
Best time for 896K FFT length: 71.241 ms.
Best time for 1024K FFT length: 79.357 ms.
Best time for 1280K FFT length: 105.720 ms.
Best time for 1536K FFT length: 124.309 ms.
Best time for 1792K FFT length: 150.861 ms.
Best time for 2048K FFT length: 167.395 ms.
optim is offline   Reply With Quote
Old 2003-11-08, 15:44   #3
optim
 
optim's Avatar
 
Nov 2003
European Union

23·13 Posts
Default

Results for the same Barton CPU overclocked with an FSB of 168MHz (DDR336) and CPU frequency at 2103MHz (the real cpu is 2600+ 1.91ghz fsb333). The memory is CL2.5.
A friend has the same configuration with FSB333 and 2088MHz Barton CPU with memory at CL2 and his result for FFT=1792 is 142.5ms.

It seems that the memory's CL timing is very important for Prime95. Am I right?

I would like someone to post benchmark for Athlon 3000+/3200+ or for Athlon64/Athlon64FX/Opteron CPUs.

AMD Athlon(tm) XP 2800+
CPU speed: 2103.29 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 29.729 ms.
Best time for 448K FFT length: 34.296 ms.
Best time for 512K FFT length: 36.470 ms.
Best time for 640K FFT length: 48.511 ms.
Best time for 768K FFT length: 57.834 ms.
Best time for 896K FFT length: 69.315 ms.
Best time for 1024K FFT length: 76.862 ms.
Best time for 1280K FFT length: 103.023 ms.
Best time for 1536K FFT length: 122.206 ms.
Best time for 1792K FFT length: 149.836 ms.
Best time for 2048K FFT length: 162.771 ms.
optim is offline   Reply With Quote
Old 2003-11-09, 07:22   #4
QuintLeo
 
QuintLeo's Avatar
 
Oct 2002
Lost in the hills of Iowa

26·7 Posts
Default

Prime is VERY memory-intensive - anything you can do to speed up the memory that does not comprimise reliability will give significant benefits on the speed of processing.
QuintLeo is offline   Reply With Quote
Old 2003-11-10, 10:33   #5
optim
 
optim's Avatar
 
Nov 2003
European Union

23×13 Posts
Default

is it okay to have a FSB333 CPU and DDR400 memory? doesn't this add latencies to the system?

I have read that a PC with FSB333 CPU and DDR333 memory is faster than a PC with FSB333 CPU and DDR400 memory due to latency issues.

Last fiddled with by optim on 2003-11-10 at 10:39
optim is offline   Reply With Quote
Old 2003-11-11, 18:33   #6
czolus
 
Sep 2003

32 Posts
Default

Yeah, it has to do with the fact that the memory controller has to insert wait states to be able to force synchronization between the memory and fsb. If they operate at the same frequency, no wait states are needed. Asynchronous hardware is non-smiley.
czolus is offline   Reply With Quote
Old 2003-11-12, 11:47   #7
optim
 
optim's Avatar
 
Nov 2003
European Union

23·13 Posts
Default

another interesting thing:

I benchmarked these two systems:

1: AMD Athlon XP 2600+ "Barton" overclocked to 2100MHz with Nvidia Nforce2 chipset

and

2: AMD Athlon XP 2600+ "Barton" overclocked to 2130MHz with VIA KT600 chipset

Both systems were using the same memory (Kingston DDR333 CL2.5)

System 1 showed an iteration time of roughly 150ms.
System 2 showed an iteration time of roughly 165ms.

there was a 15ms difference!!!

Note that system2 was overclocked 30mhz more than system1.

I suspect that the reason is the different chipset (Nvidia nForce2 vs. VIA KT600). What do you think?
optim is offline   Reply With Quote
Old 2003-11-13, 17:29   #8
E_tron
 
E_tron's Avatar
 
Sep 2002
Austin, TX

3×11×17 Posts
Default

Not the same CPU, but here are some results from a 2500+ Barton (Multiplier OC to 2800+, and FSB OC to 3200+) w/ Nforce 2 Chipset. Both setting have about the same clock speed.

Abit NF7-S rev 2.0, 2500+ Barton Retail, Zalman CNPS7000A-AlCu heatsink and fan, Cosair Value Select 512MB DDR 400

OC 2500+ Barton to 2800+ (13x166) vcore 1.7

AMD Athlon(tm)
CPU speed: 2171.27 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 29.660 ms.
Best time for 448K FFT length: 34.364 ms.
Best time for 512K FFT length: 36.730 ms.
Best time for 640K FFT length: 48.499 ms.
Best time for 768K FFT length: 58.048 ms.
Best time for 896K FFT length: 69.279 ms.
Best time for 1024K FFT length: 77.787 ms.
Best time for 1280K FFT length: 104.788 ms.
Best time for 1536K FFT length: 123.324 ms.
Best time for 1792K FFT length: 151.936 ms.
Best time for 2048K FFT length: 164.968 ms.


OC 2500+ Barton to 3200+ (200x11) vcore 1.7

AMD Athlon(tm) XP 3200+
CPU speed: 2194.11 MHz
CPU features: RDTSC, CMOV, PREFETCH, MMX, SSE
L1 cache size: 64 KB
L2 cache size: 512 KB
L1 cache line size: 64 bytes
L2 cache line size: 64 bytes
L1 TLBS: 32
L2 TLBS: 256
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 27.547 ms.
Best time for 448K FFT length: 30.966 ms.
Best time for 512K FFT length: 33.467 ms.
Best time for 640K FFT length: 45.402 ms.
Best time for 768K FFT length: 54.100 ms.
Best time for 896K FFT length: 65.043 ms.
Best time for 1024K FFT length: 71.524 ms.
Best time for 1280K FFT length: 94.931 ms.
Best time for 1536K FFT length: 112.861 ms.
Best time for 1792K FFT length: 136.426 ms.
Best time for 2048K FFT length: 150.994 ms.


It looks like bus speed helps some, but not too much.
E_tron is offline   Reply With Quote
Old 2003-11-24, 06:22   #9
E_tron
 
E_tron's Avatar
 
Sep 2002
Austin, TX

10618 Posts
Default

you could run the FSB and Memory at 200mhz (2x200=400mhz). That would speed things up. If you don’t want to overclock your actual CPU, then you can change to multiplier to something like 9.5x. That would keep the core at about 1.9ghz and the FSB at 3.2gigabits'ish.

If your CPU is a post week 39 processor, then the multiplier is most likely LOCKED and you cant do the scenario mentioned above.
E_tron 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
Does anyone have i7 920? for Benchmark? cipher Twin Prime Search 2 2009-04-14 20:16
Benchmark Weirdness R.D. Silverman Hardware 2 2007-07-25 12:16
Athlon 64 Benchmark(s) GSV3MiaC Hardware 4 2004-01-26 10:43
AMD Barton core SalemTheCat100 Hardware 13 2003-03-24 12:43

All times are UTC. The time now is 02:08.


Wed Aug 10 02:08:38 UTC 2022 up 33 days, 20:55, 1 user, load averages: 1.10, 0.98, 1.00

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 𝛴𝜎𝜍 𝛵𝜏 𝛶𝜐 𝛷𝜙𝜑 𝛸𝜒 𝛹𝜓 𝛺𝜔