Post your Prime95 Benchmarks!!

[Duvie]

I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

 

[Boogak]

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3264.42 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: 128 bytes
TLBS: 64
Prime95 version 23.8, RdtscTiming=1
Best time for 384K FFT length: 10.844 ms.
Best time for 448K FFT length: 12.885 ms.
Best time for 512K FFT length: 14.606 ms.
Best time for 640K FFT length: 17.595 ms.
Best time for 768K FFT length: 21.418 ms.
Best time for 896K FFT length: 25.588 ms.
Best time for 1024K FFT length: 28.505 ms.
Best time for 1280K FFT length: 37.292 ms.
Best time for 1536K FFT length: 45.959 ms.
Best time for 1792K FFT length: 54.644 ms.
Best time for 2048K FFT length: 61.741 ms.

 

[Duvie]

My machine is doing something worthwhiled so I will post numbers later....I am on my laptop...


Here is my 2.4@3.5 w/ 54% cpu utilization in HT with Pinnacle 8.5 encoding a DVD I authored....

I get 108.554 at the 1792k FFT length and that is on version 22.12 (faster then an athlon 3000+ )....I like this older version but I have tried the version 23.6 which had quite a nice boost as it added the 2048k reading and the 1792 reading at same speed drops like 5-10% if I remember correctly....

We all know prime is not HT enabled so we can rule that out as source of phenomenal scores...That and my 2.4b@3.24ghz would score almost identical to my 2.6c@3.25ghz....

 

[jim1976]

Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 2826.51 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: 128 bytes
TLBS: 64
Prime95 version 23.8, RdtscTiming=1
Best time for 384K FFT length: 12.666 ms.
Best time for 448K FFT length: 15.098 ms.
Best time for 512K FFT length: 17.144 ms.
Best time for 640K FFT length: 20.571 ms.
Best time for 768K FFT length: 24.998 ms.
Best time for 896K FFT length: 29.654 ms.
Best time for 1024K FFT length: 33.299 ms.
Best time for 1280K FFT length: 43.676 ms.
Best time for 1536K FFT length: 53.671 ms.
Best time for 1792K FFT length: 63.954 ms.
Best time for 2048K FFT length: 72.357 ms.

 

[Amused]

Originally posted by: Duvie
I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

My memory timings at 225/450/900 FSB are 2-2-3-7. (the 3 is RAS to CAS... there doesn't seem to be a standard on the order of these numbers) Those will probably offset the higher FSB at very loose timings just a bit, don't you think?

Plus I have PAT and Turbo enabled. So that might give me a little boost.

 

[Duvie]

Originally posted by: Amused

Originally posted by: Duvie
I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

My memory timings at 225/450/900 FSB are 2-2-3-7. (the 3 is RAS to CAS... there doesn't seem to be a standard on the order of these numbers) Those will probably offset the higher FSB at very loose timings just a bit, don't you think?

Yep!! I do believe so...I remember seeing some movement albeit not big from large ddr speed swings and cas timings....

 

[ZL1]

Originally posted by: Amused

Originally posted by: Duvie
I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

My memory timings at 225/450/900 FSB are 2-2-3-7. (the 3 is RAS to CAS... there doesn't seem to be a standard on the order of these numbers) Those will probably offset the higher FSB at very loose timings just a bit, don't you think?

ram at 1:1 helps too


D

 

[ZL1]

Originally posted by: Duvie

Originally posted by: Amused

Originally posted by: Duvie
I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

My memory timings at 225/450/900 FSB are 2-2-3-7. (the 3 is RAS to CAS... there doesn't seem to be a standard on the order of these numbers) Those will probably offset the higher FSB at very loose timings just a bit, don't you think?

Yep!! I do believe so...I remember seeing some movement albeit not big from large ddr speed swings and cas timings....

lowest cas possible to highest cas difference in Mhz is about 30-40mhz (x2) (based on my personal tests, however I usualy trust those )


D

 

[Amused]

Originally posted by: ZL1

Originally posted by: Amused

Originally posted by: Duvie
I question the whole benchmark anyways....If you think about it that 2.4@3.5 (1168fsb) should be quite sweet and with that bus could possibly negate the 3.2@3.6ghz however if you look at it the 3.5vs3.6 it is a 2.65% lead in clockspeed and the score of 58.08/56.00 is 3.7%....Hmmmm!! Pretty nice scaling.....

If you look at the crux of it we know that athlons have been notorious excellent scientific number crunchers but why so piss-poor here??? Lets look at superpi...Athlon 3000+ probably at default scores a 90sec....I get 93sec at 2.4@3.5ghz....Athlons with there high IPC and great FPU live for this stuff and it is clear the prime benchmark is another worthless cross-platform benchmark....This is way beyond intel optimised and SSE2 enhancements which AMD64 has some now....


Amused no offense to you...I think the thread was worthwhile cause it exposes this type of benchmark for what it is....QUESTIONABLE!!!!!


The benchmark is a worthwhile test when comparing on your system when ocing or tweaking and that is about it....I have seen when the system is flaky diminishing returns so it can be a sign for ocers....BUt it pretty much stops there...

My memory timings at 225/450/900 FSB are 2-2-3-7. (the 3 is RAS to CAS... there doesn't seem to be a standard on the order of these numbers) Those will probably offset the higher FSB at very loose timings just a bit, don't you think?

ram at 1:1 helps too


D

Yeah, I am at 1:1 900 FSB.

 

[Amused]

BTW, Duvie you know that Prime95 has an excellent built in torture test that will find most any instability on a system, right? It's helped me insure my systems are very stable along with SuperPi running at the same time.

 

[Duvie]

Here is a good study...It show different bus speeds, different clock speeds, different cas timings, different ratios, etc....

________________________________________________________________________________________________

-------------------------3.033ghz w/ 466ddr-------3.2ghz w/ 338ddr------3.2ghz w/ 394ddr------3.2ghz w/ 394ddr (CPC)
-----------------------233fsb,1:1,2.5,7,3,3----------250fsb,3:2,2,6,3,2----250fsb,5:4,2,6,3,3------250fsb,5:4,2,6,3,3

Sissoft MEM---------------5650's------------------------4800------------------------5380's---------------------5420's

Tmpgenc-------------------2:24---------------------------2:18------------------------2:18------------------------2:17

SuperPi 1M----------------0:45---------------------------0:44-------------------------0:43-----------------------0:43
superPI 2M---------------1:45----------------------------1:43-------------------------1:43-----------------------1:42
Prime95 (ver22)
1792k----------------------76.643------------------------73.320----------------------73.309--------------------73.137

Besweet Wav-AC3------4:25----------------------------4:11------------------------4:08------------------------4:04
DVDshrink 3.0------------17:58--------------------------17:28----------------------17:26-----------------------17:32

Cinebench2003
single cp/multi------------307/363-----------------------323/384-------------------323/381-------------------329/388
AutoCAD 2002------------0:48---------------------------0:46-------------------------0:45-----------------------0:45

WinRAR---------------------3:08---------------------------3:08------------------------3:11-------------------------3:11

3dmark2k1SE-------------9085---------------------------9076-----------------------9077-------------------------Didn't Do

_______________________________________________________________________________________________

3.033 versus 3.2ghz is 5.5% increase in clock speed

76.643 vs 73.137 is 4.5% difference in score versus the much slower ram speed but faster bus...

0.3% difference with same bus, near same timings, but different ram speed and CPC enabled in GAT

0.0x% difference with same bus, near same timings, faster ram speed, both w/o GAT

Amused your combination of the 1:1 ratio with high ddr speed, great timings made up for his 268mhz of more system bus....and left it to the fact that 3.6ghz is just faster then 3.5ghz...IN the end it still scale more on cpu speed then any other thing..

 

[Duvie]

Originally posted by: Amused
BTW, Duvie you know that Prime95 has an excellent built in torture test that will find most any instability on a system, right? It's helped me insure my systems are very stable along with SuperPi running at the same time.

Yep...been using it for 1-1/2 years now...I ran 2x instances of it for 12 hours with my current setup and I have also ran it concurrent with 6 hours tmpgenc encoding...I also like to run superpi32mb in the background of my encoding testing

 

[Amused]

Originally posted by: Duvie

Originally posted by: Amused
BTW, Duvie you know that Prime95 has an excellent built in torture test that will find most any instability on a system, right? It's helped me insure my systems are very stable along with SuperPi running at the same time.

Yep...been using it for 1-1/2 years now...I ran 2x instances of it for 12 hours with my current setup and I have also ran it concurrent with 6 hours tmpgenc encoding...I also like to run superpi32mb in the background of my encoding testing

Cool! Yeah, I've found running it with SuperPi will weed out any instability faster than just running Prime alone. Plus I like to open, use and close random programs while running them. Just to see if anything locks or craps out.

A fun thing to do is run Prime blended TT with SuperPi and try surfing the web. Feels like an old 386

 

[Duvie]

Originally posted by: Don66
Mine here.

Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 3507.95 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.214 ms.
Best time for 448K FFT length: 12.165 ms.
Best time for 512K FFT length: 13.825 ms.
Best time for 640K FFT length: 16.513 ms.
Best time for 768K FFT length: 20.153 ms.
Best time for 896K FFT length: 23.852 ms.
Best time for 1024K FFT length: 26.723 ms.
Best time for 1280K FFT length: 35.116 ms.
Best time for 1536K FFT length: 43.218 ms.
Best time for 1792K FFT length: 51.293 ms.
Best time for 2048K FFT length: 58.080 ms.
[Tue Dec 23 18:35:49 2003]

Here is my 23.7 score and since I am at same speed but 168mhz more fsb I wonder if there are any ram differences as well....

I am running currently 3:2 (390ddr) cas 2,6,3,2 Gat CPC enabled...

[Tue Dec 23 22:26:42 2003]
Compare your results to other computers at http://www.mersenne.org/bench.htm
That web page also contains instructions on how your results can be included.

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.93 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.179 ms.
Best time for 448K FFT length: 12.105 ms.
Best time for 512K FFT length: 13.780 ms.
Best time for 640K FFT length: 16.466 ms.
Best time for 768K FFT length: 19.994 ms.
Best time for 896K FFT length: 23.740 ms.
Best time for 1024K FFT length: 26.627 ms.
Best time for 1280K FFT length: 34.924 ms.
Best time for 1536K FFT length: 42.987 ms.
Best time for 1792K FFT length: 51.004 ms.
Best time for 2048K FFT length: 57.737 ms.

A little more studying...

Reran 3 times difference was +/- .03 max...very consistent. An obvious upgrade from past 3dmark test....

lowest at same background settings...

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3508.02 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.178 ms.
Best time for 448K FFT length: 12.089 ms.
Best time for 512K FFT length: 13.791 ms.
Best time for 640K FFT length: 16.470 ms.
Best time for 768K FFT length: 20.001 ms.
Best time for 896K FFT length: 23.733 ms.
Best time for 1024K FFT length: 26.623 ms.
Best time for 1280K FFT length: 34.924 ms.
Best time for 1536K FFT length: 43.025 ms.
Best time for 1792K FFT length: 51.030 ms.
Best time for 2048K FFT length: 57.706 ms.

Here it is with shutdown of all background apps...not a reboot...

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3508.05 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.179 ms.
Best time for 448K FFT length: 12.096 ms.
Best time for 512K FFT length: 13.757 ms.
Best time for 640K FFT length: 16.461 ms.
Best time for 768K FFT length: 19.995 ms.
Best time for 896K FFT length: 23.741 ms.
Best time for 1024K FFT length: 26.596 ms.
Best time for 1280K FFT length: 34.935 ms.
Best time for 1536K FFT length: 42.957 ms.
Best time for 1792K FFT length: 51.011 ms.
Best time for 2048K FFT length: 57.683 ms.
[Tue Dec 23 22:38:01 2003]


here it is with a fresh reboot and background apps shutdown....

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.94 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.176 ms.
Best time for 448K FFT length: 12.079 ms.
Best time for 512K FFT length: 13.753 ms.
Best time for 640K FFT length: 16.467 ms.
Best time for 768K FFT length: 19.988 ms.
Best time for 896K FFT length: 23.715 ms.
Best time for 1024K FFT length: 26.600 ms.
Best time for 1280K FFT length: 34.919 ms.
Best time for 1536K FFT length: 42.977 ms.
Best time for 1792K FFT length: 51.069 ms.
Best time for 2048K FFT length: 57.576 ms.

Gotta love these test.....overall only a 0.28% increase

reran 6 times afterwards and numbers went as low as 57.549 to 57.594

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.94 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.164 ms.
Best time for 448K FFT length: 12.081 ms.
Best time for 512K FFT length: 13.759 ms.
Best time for 640K FFT length: 16.452 ms.
Best time for 768K FFT length: 19.985 ms.
Best time for 896K FFT length: 23.739 ms.
Best time for 1024K FFT length: 26.592 ms.
Best time for 1280K FFT length: 34.921 ms.
Best time for 1536K FFT length: 43.011 ms.
Best time for 1792K FFT length: 51.030 ms.
Best time for 2048K FFT length: 57.658 ms.


For what it is worth...

 

[BlvdKing]

AMD Athlon XP 2500+ (Barton)

AMD Athlon(tm)
CPU speed: 1834.17 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 22.12, RdtscTiming=1
Best time for 256K FFT length: 19.834 ms.
Best time for 320K FFT length: 25.627 ms.
Best time for 384K FFT length: 32.528 ms.
Best time for 448K FFT length: 36.706 ms.
Best time for 512K FFT length: 40.257 ms.
Best time for 640K FFT length: 52.578 ms.
Best time for 768K FFT length: 64.180 ms.
Best time for 896K FFT length: 75.606 ms.
Best time for 1024K FFT length: 85.736 ms.
Best time for 1280K FFT length: 113.230 ms.
Best time for 1536K FFT length: 136.339 ms.
Best time for 1792K FFT length: 164.195 ms.

Something is wierd on my machine. Prime95 flucuates from 3% to 100% CPU constantly during torture test, but it doesn't stay at 100% for more than a second, no matter what priority it's at. Something must be eating CPU cycles somewhere but I don't see anything in the task manager that is suspicious.

 

[Amused]

Originally posted by: Duvie

Originally posted by: Don66
Mine here.

Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 3507.95 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.214 ms.
Best time for 448K FFT length: 12.165 ms.
Best time for 512K FFT length: 13.825 ms.
Best time for 640K FFT length: 16.513 ms.
Best time for 768K FFT length: 20.153 ms.
Best time for 896K FFT length: 23.852 ms.
Best time for 1024K FFT length: 26.723 ms.
Best time for 1280K FFT length: 35.116 ms.
Best time for 1536K FFT length: 43.218 ms.
Best time for 1792K FFT length: 51.293 ms.
Best time for 2048K FFT length: 58.080 ms.
[Tue Dec 23 18:35:49 2003]

Here is my 23.7 score and since I am at same speed but 168mhz more fsb I wonder if there are any ram differences as well....

I am running currently 3:2 (390ddr) cas 2,6,3,2 Gat CPC enabled...

[Tue Dec 23 22:26:42 2003]
Compare your results to other computers at http://www.mersenne.org/bench.htm
That web page also contains instructions on how your results can be included.

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.93 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.179 ms.
Best time for 448K FFT length: 12.105 ms.
Best time for 512K FFT length: 13.780 ms.
Best time for 640K FFT length: 16.466 ms.
Best time for 768K FFT length: 19.994 ms.
Best time for 896K FFT length: 23.740 ms.
Best time for 1024K FFT length: 26.627 ms.
Best time for 1280K FFT length: 34.924 ms.
Best time for 1536K FFT length: 42.987 ms.
Best time for 1792K FFT length: 51.004 ms.
Best time for 2048K FFT length: 57.737 ms.

A little more studying...

Reran 3 times difference was +/- .03 max...very consistent. An obvious upgrade from past 3dmark test....

lowest at same background settings...

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3508.02 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.178 ms.
Best time for 448K FFT length: 12.089 ms.
Best time for 512K FFT length: 13.791 ms.
Best time for 640K FFT length: 16.470 ms.
Best time for 768K FFT length: 20.001 ms.
Best time for 896K FFT length: 23.733 ms.
Best time for 1024K FFT length: 26.623 ms.
Best time for 1280K FFT length: 34.924 ms.
Best time for 1536K FFT length: 43.025 ms.
Best time for 1792K FFT length: 51.030 ms.
Best time for 2048K FFT length: 57.706 ms.

Here it is with shutdown of all background apps...not a reboot...

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3508.05 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.179 ms.
Best time for 448K FFT length: 12.096 ms.
Best time for 512K FFT length: 13.757 ms.
Best time for 640K FFT length: 16.461 ms.
Best time for 768K FFT length: 19.995 ms.
Best time for 896K FFT length: 23.741 ms.
Best time for 1024K FFT length: 26.596 ms.
Best time for 1280K FFT length: 34.935 ms.
Best time for 1536K FFT length: 42.957 ms.
Best time for 1792K FFT length: 51.011 ms.
Best time for 2048K FFT length: 57.683 ms.
[Tue Dec 23 22:38:01 2003]


here it is with a fresh reboot and background apps shutdown....

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.94 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.176 ms.
Best time for 448K FFT length: 12.079 ms.
Best time for 512K FFT length: 13.753 ms.
Best time for 640K FFT length: 16.467 ms.
Best time for 768K FFT length: 19.988 ms.
Best time for 896K FFT length: 23.715 ms.
Best time for 1024K FFT length: 26.600 ms.
Best time for 1280K FFT length: 34.919 ms.
Best time for 1536K FFT length: 42.977 ms.
Best time for 1792K FFT length: 51.069 ms.
Best time for 2048K FFT length: 57.576 ms.

Gotta love these test.....overall only a 0.28% increase

rerab 3 times afterwards and numbers went back up to one without reboot but background items shutdown....

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.94 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.164 ms.
Best time for 448K FFT length: 12.081 ms.
Best time for 512K FFT length: 13.759 ms.
Best time for 640K FFT length: 16.452 ms.
Best time for 768K FFT length: 19.985 ms.
Best time for 896K FFT length: 23.739 ms.
Best time for 1024K FFT length: 26.592 ms.
Best time for 1280K FFT length: 34.921 ms.
Best time for 1536K FFT length: 43.011 ms.
Best time for 1792K FFT length: 51.030 ms.
Best time for 2048K FFT length: 57.658 ms.


For what it is worth...

Wow, you went nuts. Great numbers, though! For the record, my test was with a fresh boot and no background aps running as well. My XP install is only a week or so old, too. It seems Prime95 is mostly dependent on processor speed, as your scores are just under mine, and your CPU speed is just 100MHz shy of mine.

What FSB, ratio and timings are you running right now?

 

[rb56]

Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 2798.61 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 12.709 ms.
Best time for 448K FFT length: 15.097 ms.
Best time for 512K FFT length: 17.214 ms.
Best time for 640K FFT length: 20.579 ms.
Best time for 768K FFT length: 25.038 ms.
Best time for 896K FFT length: 29.692 ms.
Best time for 1024K FFT length: 33.349 ms.
Best time for 1280K FFT length: 43.753 ms.
Best time for 1536K FFT length: 53.753 ms.
Best time for 1792K FFT length: 63.939 ms.
Best time for 2048K FFT length: 72.341 ms.

 

[Duvie]

I am running 292fsb w/ 1.58v vcore 3:2 ratio cas 2,6,3,2 CPC enabled.... I get 5791/5784 in snadra mem scores for what that is worth...10480 in pcmark2002 memory as well


I think the barton 2500+ score looks in line compared to the 3200+ up above when you ratio the real clock speed...

One thing though mine held 50% which for HT enabled chip is 100%....So I am not sure what is going on...Where you using the task manager or some other program that may refresh those numbers faster???

One thing for sure the prime program like the 512kb of cache as seen from the 2100+xp numbers to your barton 2500+ numbers which represents only a 100mhz real clock speed difference....But the score is 50% better....

I ran 2 instances of the bench but since it is impossible to start them at same time lets look at the 1792k reading...

I get 78.306ms in one and 78.085 in the other .....regularly I get 51ms so it appears HT may be helping it with 2 instances as it did not double but in fact about 20% faster...I will turn off bios HT and retest...

God I have too much time on my hands!!!!

 

[Amused]

Originally posted by: rb56
Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 2798.61 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 12.709 ms.
Best time for 448K FFT length: 15.097 ms.
Best time for 512K FFT length: 17.214 ms.
Best time for 640K FFT length: 20.579 ms.
Best time for 768K FFT length: 25.038 ms.
Best time for 896K FFT length: 29.692 ms.
Best time for 1024K FFT length: 33.349 ms.
Best time for 1280K FFT length: 43.753 ms.
Best time for 1536K FFT length: 53.753 ms.
Best time for 1792K FFT length: 63.939 ms.
Best time for 2048K FFT length: 72.341 ms.

Those are some great numbers for a 2.8 @2.8. I wonder what they would look like if you OCed to 3.2 or above? Have you tried it?

Heh, now that my machine is done, I wanna OC someone else's

 

[Amused]

Originally posted by: Duvie
I am running 292fsb w/ 1.58v vcore 3:2 ratio cas 2,6,3,2 CPC enabled.... I get 5791/5784 in snadra mem scores for what that is worth...10480 in pcmark2002 memory as well


I think the barton 2500+ score looks in line compared to the 3200+ up above when you ratio the real clock speed...

One thing though mine held 50% which for HT enabled chip is 100%....So I am not sure what is going on...Where you using the task manager or some other program that may refresh those numbers faster???

One thing for sure the prime program like the 512kb of cache as seen from the 2100+xp numbers to your barton 2500+ numbers which represents only a 100mhz real clock speed difference....But the score is 50% better....

I ran 2 instances of the bench but since it is impossible to start them at same time lets look at the 1792k reading...

I get 78.306ms in one and 78.085 in the other .....regularly I get 51ms so it appears HT may be helping it with 2 instances as it did not double but in fact about 20% faster...I will turn off bios HT and retest...

God I have too much time on my hands!!!!

Yeah, but it's fun, isn't it?

By the way, my brain is fried, what is your memory speed after the ratio x2? Like mine is 1:1 225 so it's 450. What is yours?

 

[Duvie]

Interesting results with HT off in Bios...

Again impossible to start at same time, but overall weird...

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.93 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.133 ms.
Best time for 448K FFT length: 12.004 ms.
Best time for 512K FFT length: 13.701 ms.
Best time for 640K FFT length: 16.333 ms.
Best time for 768K FFT length: 19.941 ms.
Best time for 896K FFT length: 23.614 ms.
Best time for 1024K FFT length: 26.524 ms.
Best time for 1280K FFT length: 66.098 ms.
Best time for 1536K FFT length: 74.254 ms.
Best time for 1792K FFT length: 82.029 ms.
Best time for 2048K FFT length: 88.864 ms.

Intel(R) Pentium(R) 4 CPU 2.40GHz
CPU speed: 3507.90 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 10.134 ms.
Best time for 448K FFT length: 12.004 ms.
Best time for 512K FFT length: 13.697 ms.
Best time for 640K FFT length: 16.350 ms.
Best time for 768K FFT length: 19.947 ms.
Best time for 896K FFT length: 23.613 ms.
Best time for 1024K FFT length: 26.537 ms.
Best time for 1280K FFT length: 50.675 ms.
Best time for 1536K FFT length: 74.182 ms.
Best time for 1792K FFT length: 82.051 ms.
Best time for 2048K FFT length: 57.456 ms.

notice how up until 1280k the test ran themabout what HT on did...However it was clear the program was swapping back and forth in terms of priority of use....

At 1536 numbers got big as if they were competing with each other...at 1792 we see a combined 164 versus 156 so only a 5% gain....

However to top it with the cherry my best single test score at 57.456....

Reran single test and I get 57.543 as a single app....

 

[rb56]

Originally posted by: Amused

Originally posted by: rb56
Intel(R) Pentium(R) 4 CPU 2.80GHz
CPU speed: 2798.61 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: 128 bytes
TLBS: 64
Prime95 version 23.7, RdtscTiming=1
Best time for 384K FFT length: 12.709 ms.
Best time for 448K FFT length: 15.097 ms.
Best time for 512K FFT length: 17.214 ms.
Best time for 640K FFT length: 20.579 ms.
Best time for 768K FFT length: 25.038 ms.
Best time for 896K FFT length: 29.692 ms.
Best time for 1024K FFT length: 33.349 ms.
Best time for 1280K FFT length: 43.753 ms.
Best time for 1536K FFT length: 53.753 ms.
Best time for 1792K FFT length: 63.939 ms.
Best time for 2048K FFT length: 72.341 ms.

Those are some great numbers for a 2.8 @2.8. I wonder what they would look like if you OCed to 3.2 or above? Have you tried it?

Heh, now that my machine is done, I wanna OC someone else's

Not yet, I set it up a while back but have been working way too much to play with it. But I've got the rest of the year off so after I set up raid maybe a little OC will be in order.

rb56

 

[BlvdKing]

I run the task manager while running the P95 torture test. The page file usage is cyclical and very predictable. I disabled some stuff in msconfig that I did not know was enabled (like CTHelper) so I'll run it again overnight and see what happens.

BTW, that HT is great stuff

 

[Duvie]

Originally posted by: Amused

Originally posted by: Duvie
I am running 292fsb w/ 1.58v vcore 3:2 ratio cas 2,6,3,2 CPC enabled.... I get 5791/5784 in snadra mem scores for what that is worth...10480 in pcmark2002 memory as well


I think the barton 2500+ score looks in line compared to the 3200+ up above when you ratio the real clock speed...

One thing though mine held 50% which for HT enabled chip is 100%....So I am not sure what is going on...Where you using the task manager or some other program that may refresh those numbers faster???

One thing for sure the prime program like the 512kb of cache as seen from the 2100+xp numbers to your barton 2500+ numbers which represents only a 100mhz real clock speed difference....But the score is 50% better....

I ran 2 instances of the bench but since it is impossible to start them at same time lets look at the 1792k reading...

I get 78.306ms in one and 78.085 in the other .....regularly I get 51ms so it appears HT may be helping it with 2 instances as it did not double but in fact about 20% faster...I will turn off bios HT and retest...

God I have too much time on my hands!!!!

Yeah, but it's fun, isn't it?

By the way, my brain is fried, what is your memory speed after the ratio x2? Like mine is 1:1 225 so it's 450. What is yours?

It is 292/3 = 194.6666x2= 389.3333ddr....

I wish I coud do more but it appears my Geil Ultras that will do 466mhz ddr at 233fsb may have the CH-5 chips and hence past 255fsb no ratio works except 3:2....I only have 2.8v of vdimm options and ofourse my abit undervolts that to 2.78v...I could probably be helped with some more vdimm but how much I don't know....

I ve done the testing and high ddr speed unless i can keep cas 2 timings isn't worth it anyways...

 

[Amused]

Originally posted by: Duvie

Originally posted by: Amused

Originally posted by: Duvie
I am running 292fsb w/ 1.58v vcore 3:2 ratio cas 2,6,3,2 CPC enabled.... I get 5791/5784 in snadra mem scores for what that is worth...10480 in pcmark2002 memory as well


I think the barton 2500+ score looks in line compared to the 3200+ up above when you ratio the real clock speed...

One thing though mine held 50% which for HT enabled chip is 100%....So I am not sure what is going on...Where you using the task manager or some other program that may refresh those numbers faster???

One thing for sure the prime program like the 512kb of cache as seen from the 2100+xp numbers to your barton 2500+ numbers which represents only a 100mhz real clock speed difference....But the score is 50% better....

I ran 2 instances of the bench but since it is impossible to start them at same time lets look at the 1792k reading...

I get 78.306ms in one and 78.085 in the other .....regularly I get 51ms so it appears HT may be helping it with 2 instances as it did not double but in fact about 20% faster...I will turn off bios HT and retest...

God I have too much time on my hands!!!!

Yeah, but it's fun, isn't it?

By the way, my brain is fried, what is your memory speed after the ratio x2? Like mine is 1:1 225 so it's 450. What is yours?

It is 292/3 = 194.6666x2= 389.3333ddr....

I wish I coud do more but it appears my Geil Ultras that will do 466mhz ddr at 233fsb may have the CH-5 chips and hence past 255fsb no ratio works except 3:2....I only have 2.8v of vdimm options and ofourse my abit undervolts that to 2.78v...I could probably be helped with some more vdimm but how much I don't know....

I ve done the testing and high ddr speed unless i can keep cas 2 timings isn't worth it anyways...

What mobo are you using? Isn't there a vdimm mod for some of them?