DDR490 TwinMos - 2.6C ***WOW*** (P4G8X vs P4C800) SpecViewPerf 7.0 Benched

[jhites]

** Updated to include Kingston HyperX-PC3500 Ram and
TwinMos PC3700 1Gb (2X512Mb) **
SpecViewPerf 7.0; Prime95 Bench; PIFast are @
270fsb=3515Mhz @ (5:4) DDR433 CL2-3-3-6

SYSTEM

. . 2.53 . SL6S2 - - L249 Malay
. . 2.60 . SL6WH - - L311 Malay

Asus P4G8X Deluxe - BIOS ver.1004-Tweak
Asus P4C800 Deluxe ? BIOS ver.1006
Corsair TwinX512-3200LL 2X256mb
ATI AIW Radeon 9700 128Mb (Default Settings)
Saphire Radeon 9000 64Mb (Default Settings) 2.60C on P4G8X Only
Memory Timing Set CL2-6-2-2-1T (Idle Timer=Infinite)
PCI locked at 66/33[/b]

P4G8X-D vcore = 1.525v BIOS; 1.552v Actual
P4C800D vcore = 1.575v BIOS; 1.554v Actual
2.53b tests at 179fsb = 3401Mhz
P4C800D at 270fsb = 1.675v BIOS; 1.700v Actual

SpecViewPerf 7.0

---------- SUM_RESULTS\3DSMAX\SUMMARY.TXT
3dsmax-01 Weighted Geometric Mean = 11.95

---------- SUM_RESULTS\DRV\SUMMARY.TXT
drv-08 Weighted Geometric Mean = 43.72

---------- SUM_RESULTS\DX\SUMMARY.TXT
dx-07 Weighted Geometric Mean = 66.06

---------- SUM_RESULTS\LIGHT\SUMMARY.TXT
light-05 Weighted Geometric Mean = 15.53

---------- SUM_RESULTS\PROE\SUMMARY.TXT
proe-01 Weighted Geometric Mean = 15.82

---------- SUM_RESULTS\UGS\SUMMARY.TXT
ugs-01 Weighted Geometric Mean = 13.10

PiFast ver4.2

Program : PiFast version 4.2, by Xavier Gourdon
Computation of 10000000 digits of Pi
Method used : Chudnovsky
Size of FFT : 1024 K
Physical memory used : ~ 61355 K
Disk memory used : ~ 0.00 Meg
------------------------------------------------------------
Computation run information :

Start : Tue May 13 00:36:39 2003
End : Tue May 13 00:37:21 2003
Duration : 41.86 seconds
========================================
Total computation time : 41.86 seconds (~ 0.01 hours)
========================================

Seti Bench - 2 hours 10.8 minutes

minor_version=8
platform=i386-winnt-cmdline
cpu_type=Intel Pentium
system=Windows NT: 5.1
id=3118e812
key0=554d2dae
key1=3eefcafe
end_request_header
type=result
task=seti
version=256
name=21se00aa.1002.33314.573572.103
cpu_time=7856.796875
status=0
params_index=0

Two instances of Seti Bench at the same time

cpu_time=10717.125000
cpu_time=10716.937500

Prime95 Bench

Intel(R) Pentium(R) 4 CPU 2.60GHz
CPU speed: 3515.06 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: 6.805 ms.
Best time for 320K FFT length: 9.040 ms.
Best time for 384K FFT length: 10.931 ms.
Best time for 448K FFT length: 13.154 ms.
Best time for 512K FFT length: 14.807 ms.
Best time for 640K FFT length: 19.204 ms.
Best time for 768K FFT length: 23.440 ms.
Best time for 896K FFT length: 28.817 ms.
Best time for 1024K FFT length: 31.091 ms.
Best time for 1280K FFT length: 43.891 ms.
Best time for 1536K FFT length: 54.740 ms.
Best time for 1792K FFT length: 66.943 ms.

SiSoft Sandra 2003

Memory Buffered
P4G8X - 2.53 = . . . 4470 / 4441 . . DDR360 . . . . . 179fsb . . . . 3401Mhz
P4G8X - 2.60 = . . . 4395 / 4399 . . DDR360 . . . . . 179fsb . . . . 2327Mhz
P4G8X - 2.60 = . . . 4921 / 4923 . . DDR402 . . . . . 201fsb . . . . 2613Mhz
P4C800-2.53 = . . . 4417 / 4415 . . DDR360 . . . . . 179fsb . . . . 3401Mhz
P4C800-2.60 = . . . 4654 / 4659 . . DDR400(1:1) . 200fsb . . . . 2600Mhz
P4C800-2.60 = . . . 5421 / 5387 . . DDR400(5:4) . 250fsb . . . . 3250Mhz
P4C800-2.60 = . . . 5622 / 5621 . . DDR416(5:4) . 260fsb . . . . 3380Mhz
P4C800-2.60 = . . . 5271 / 5265 . . DDR360(3:2) . 270fsb . . . . 3510Mhz
****Kingston HyperX KH3500/256 256MB DDR433 PC3500****
P4C800-2.60 = . . . 4662 / 4639 . . DDR400(1:1) . 200fsb . . . . 2600Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 5421 / 5402 . . DDR400(5:4) . 250fsb . . . . 3250Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 5619 / 5627 . . DDR416(5:4) . 260fsb . . . . 3380Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 5757 / 5727 . . DDR432(5:4) . 270fsb . . . . 3510Mhz (2.0-3-3-6)
TwinMos PC3700 DDR466 @ 2.5-3-3-7
P4C800-2.60 = . . . 5310 / 5307 . . DDR466(1:1) . 233fsb . . . . 3029Mhz
P4C800-2.60 = . . . 5487 / 5462 . . DDR480(1:1) . 240fsb . . . . 3121Mhz
P4C800-2.60 = . . . 5600 / 5576 . . DDR490(1:1) . 245fsb . . . . 3185Mhz
P4C800-2.60 = . . . 5676 / 5639 . . DDR492(1:1) . 246fsb . . . . 3198Mhz
P4C800-2.60 = . . . 5745 / 5727 . . DDR432(5:4) . 270fsb . . . . 3510Mhz (2.0-3-2-6)


Memory UnBuffered
P4G8X - 2.53 = . . . 2104 / 2120 . . DDR360 . . . . . 179fsb . . . . 3401Mhz
P4G8X - 2.60 = . . . 2093 / 2135 . . DDR360 . . . . . 179fsb . . . . 2327Mhz
P4G8X - 2.60 = . . . 2314 / 2358 . . DDR402 . . . . . 201fsb . . . . 2613Mhz
P4C800-2.53 = . . . 2494 / 2537 . . DDR360 . . . . . 179fsb . . . . 3401Mhz
P4C800-2.60 = . . . 2650 / 2731 . . DDR400(1:1) . 200fsb . . . . 2600Mhz
P4C800-2.60 = . . . 2684 / 2745 . . DDR400(5:4) . 250fsb . . . . 3250Mhz
P4C800-2.60 = . . . 2804 / 2872 . . DDR400(5:4) . 260fsb . . . . 3380Mhz
P4C800-2.60 = . . . 2500 / 2541 . . DDR360(3:2) . 270fsb . . . . 3510Mhz
****Kingston HyperX KH3500/256 256MB DDR433 PC3500****
P4C800-2.60 = . . . 2637 / 2720 . . DDR400(1:1) . 200fsb . . . . 2600Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 2682 / 2741 . . DDR400(5:4) . 250fsb . . . . 3250Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 2806 / 2862 . . DDR416(5:4) . 260fsb . . . . 3380Mhz (2.0-2-2-6)
P4C800-2.60 = . . . 2650 / 2672 . . DDR432(5:4) . 270fsb . . . . 3510Mhz (2.0-3-3-6)
TwinMos PC3700 DDR466 @ 2.5-3-3-7
P4C800-2.60 = . . . 3085 / 3109 . . DDR466(1:1) . 233fsb . . . . 3029Mhz
P4C800-2.60 = . . . 3101 / 3173 . . DDR480(1:1) . 240fsb . . . . 3121Mhz
P4C800-2.60 = . . . 3155 / 3278 . . DDR490(1:1) . 245fsb . . . . 3185Mhz
P4C800-2.60 = . . . 3166 / 3251 . . DDR492(1:1) . 246fsb . . . . 3198Mhz
P4C800-2.60 = . . . 2941 / 3007 . . DDR432(5:4) . 270fsb . . . . 3510Mhz (2.0-3-2-6)


CPU Arithmetic . . . . . . . . . . . . . . . . . . . . . . FSB
P4G8X-2.53 = . . . 8862/1974/4486 . . . . 179
P4G8X-2.60 = . . . 6889/2025/4534 . . . . 179
P4G8X-2.60 = . . . 7822/2261/5036 . . . . 201
P4C800-2.53 = . . 8966/1970/4502 . . . . 179
P4C800-2.60 = . . 7725/2252/5000 . . . . 200
P4C800-2.60 = . 10070/2808/6226 . . . . 250
P4C800-2.60 = . 10331/2917/6482 . . . . 260
P4C800-2.60 = . 10846/3033/6475 . . . . 270


CPU Multi-Media . . . . . . . . . . . . . . . . . . . . FSB . . . Radeon
P4G8X - 2.53 = . . . 13654 / 17346 . . . . 179 . . . . 9700
P4G8X - 2.60 = . . . 10811 / 17156 . . . . 179 . . . . 9000
P4G8X - 2.60 = . . . 12076 / 19186 . . . . 201 . . . . 9000
P4C800-2.53= . . . 13575 / 17343 . . . . 179 . . . . 9700
P4C800-2.60= . . . 12044 / 19103 . . . . 200 . . . . 9700
P4C800-2.60= . . . 15008 / 23301 . . . . 250 . . . . 9700
P4C800-2.60= . . . 15617 / 24659 . . . . 260 . . . . 9700
P4C800-2.60= . . . 16224 / 25606 . . . . 270 . . . . 9700


PC Mark 2002
. . . . . . . . . . . . . . . . . CPU . . . . . . . . Memory . . . . HDD . . . . FSB
P4G8X - 2.53 = . . . 8368 . . . . . . 9061 (1:1) . . . 1194 . . . . 179
P4G8X - 2.60 = . . . 5802 . . . . . . 7314 (1:1) . . . 0843 . . . . 179
P4G8X - 2.60 = . . . 6475 . . . . . . 8053 (1:1) . . . 0843 . . . . 201
P4C800-2.53 = . . . 8455 . . . . . . 9268 (1:1) . . . 1144 . . . . 179
P4C800-2.60 = . . . 6408 . . . . . . 8392 (1:1) . . . 1266 . . . . 200 . . . DDR400
P4C800-2.60 = . . . 7991 . . . . . . 9521 (5:4) . . . 1258 . . . . 250 . . . DDR400
P4C800-2.60 = . . . 8271 . . . . . . 9902 (5:4) . . . 1249 . . . . 260 . . . DDR416
P4C800-2.60 = . . . 8691 . . . . . . 9808(3:2) . . . 1240 . . . . 270 . . . DDR360
****Kingston HyperX KH3500/256 256MB DDR433 PC3500****
P4C800-2.60 = . . . 6449 . . . . . . 8479(1:1) . . . 1206 . . . . 200 . . . DDR400 (2.0-2-2-6)
P4C800-2.60 = . . . 8066 . . . . . . 9592(5:4) . . . 1206 . . . . 250 . . . DDR400 (2.0-2-2-6)
P4C800-2.60 = . . . 8400 . . . . . 10043(5:4) . . . 1193 . . . . 260 . . . DDR416 (2.0-2-2-6)
P4C800-2.60 = . . . 8702 . . . . . 10142(5:4) . . . 1191 . . . . 270 . . . DDR432 (2.0-3-3-6) PCMark2002
TwinMos PC3700 DDR466 @ 2.5-3-3-7
P4C800-2.60 = . . . 7882 . . . . . 10462(1:1) . . . 1146 . . . . 245 . . . DDR490 (2.5-3-3-7)
P4C800-2.60 = . . . 8674 . . . . . 10378(5:4) . . . 1153 . . . . 270 . . . DDR432 (2.0-3-2-6)


3D Mark 2001se
P4G8X ? 2.53 . . . . 17386 . . Radeon 9700 . . . 179fsb(1:1) . . . . 3401Mhz
P4G8X ? 2.60 . . . . 04883 . . Radeon 9000 . . . 179fsb
P4G8X ? 2.60 . . . . 04891 . . Radeon 9000 . . . 201fsb
P4C800?2.53 . . . . 17458 . . Radeon 9700 . . . 179fsb(1:1) . . . . 3401Mhz
P4C800?2.60 . . . . 15503 . . Radeon 9700 . . . 200fsb(1:1) . . . . 2600Mhz
P4C800?2.60 . . . . 16878 . . Radeon 9700 . . . 250fsb(5:4) . . . . 3250Mhz
P4C800?2.60 . . . . 17905 . . Radeon 9700 . . . 260fsb(5:4) . . . . 3380Mhz
P4C800?2.60 . . . . 17928 . . Radeon 9700 . . . 270fsb(3:2) . . . . 3510Mhz
****Kingston HyperX KH3500/256 256MB DDR433 PC3500****
P4C800?2.60 . . . . 18005 . . Radeon 9700 . . . 270fsb(5:4) . . . . 3510Mhz 3DMark2001SE

 

[mechBgon]

I'd be interested in seeing someone do the UT/WinZip benchie with the P4C800 (and the P4G8X too). Procedure described in my second post in this thread.

 

[Thor86]

So I'm assuming the 2.6C works with the P4G8X? Great stuff.

 

[THUGSROOK]

Originally posted by: mechBgon
I'd be interested in seeing someone do the UT/WinZip benchie with the P4C800 (and the P4G8X too). Procedure described in my second post in this thread.

mechBgon ~ i just wanted to say i 100% fully endorse that test (winzip should have a timer built in for benching) however i dont have a UT cd

is there anyway you could also start a serious sam CD or q3 CD zip test?


jhites ~ im a little confused about the MHz and FSb speeds on those tests. could you clarify a little more?

 

[mechBgon]

Originally posted by: THUGSROOK

Originally posted by: mechBgon
I'd be interested in seeing someone do the UT/WinZip benchie with the P4C800 (and the P4G8X too). Procedure described in my second post in this thread.

mechBgon ~ i just wanted to say i 100% fully endorse that test (winzip should have a timer built in for benching) however i dont have a UT cd

is there anyway you could also start a serious sam CD or q3 CD zip test?


jhites ~ im a little confused about the MHz and FSb speeds on those tests. could you clarify a little more?

It doesn't have to be UT but it was about the only game I had that was popular (a Descent3/WinZip benchie works for you and I, but everyone else plays those wimpy FPS games where you stick to the floor ). jbond04's result with i850E/RDRAM remains the high-water mark (in this particular benchmark), and I keep wanting to see what other new Intel platforms can do to approach or surpass that.

 

[SupermanCK]

jhites ~ im a little confused about the MHz and FSb speeds on those tests. could you clarify a little more?

 

[dbwillis]

3D Mark 2001se

P4G8X ? 2.53 . . . . 17386 . . Radeon 9700 . . . 179fsb
P4G8X ? 2.60 . . . . 04883 . . Radeon 9000 . . . 179fsb
P4G8X ? 2.60 . . . . 04891 . . Radeon 9700 . . . 201fsb
P4C800?2.53 . . . . 17458 . . Radeon 9700 . . . 179fsb

IS that a typo?

should be Radeon 9700? not 9000?
Whys the score way down on the 2.60?

 

[jhites]

Originally posted by: THUGSROOK


mechBgon ~ i just wanted to say i 100% fully endorse that test (winzip should have a timer built in for benching) however i dont have a UT cd

is there anyway you could also start a serious sam CD or q3 CD zip test?

jhites ~ im a little confused about the MHz and FSb speeds on those tests. could you clarify a little more?

mechBgon - I will be glad to do that but it may be a couple of days.
I have not loaded much on my PC at this point. Just got MS Office on it today.
I am trying to compare with some stats that I already had from my 2.53 and boards.

Thugs - Did me adding the fsb and chip speed to some benches clear the confusion?

 

[jhites]

Originally posted by: dbwillis
3D Mark 2001se

P4G8X ? 2.53 . . . . 17386 . . Radeon 9700 . . . 179fsb
P4G8X ? 2.60 . . . . 04883 . . Radeon 9000 . . . 179fsb
P4G8X ? 2.60 . . . . 04891 . . Radeon 9700 . . . 201fsb
P4C800?2.53 . . . . 17458 . . Radeon 9700 . . . 179fsb

IS that a typo?

should be Radeon 9700? not 9000?
Whys the score way down on the 2.60?

I put the P4G8X in my wifes PC and she only has the
Radeon 64Mb 9000. This was in the first sentence of the post.

 

[jhites]

The best that I could get out of the P4G8X was 203fsb with the 2.6C
Not sure yet if it is a motherboard limitation or my Corsair PC3200LL
but the next step is 260fsb on this P4C800 and that might indicate something.

 

[THUGSROOK]

much better ~ TY

the C chip hasnt beatin the B chip in 3dmark yet? .......

 

[jhites]

Originally posted by: THUGSROOK
much better ~ TY

the C chip hasnt beatin the B chip in 3dmark yet? .......

OH but it will in just a minute.

Looks like it is a board limitation on the P4G8X fsb.
The ram is running DDR416 @ 2-3-2-6-1T (Needed 2.75v to get there though)

 

[Ardy]

Do you have hyper-threading enabled on the P4G8X ?

 

[jhites]

Originally posted by: Ardy
Do you have hyper-threading enabled on the P4G8X ?

Yes it was enabled.

EDIT
Had to go to 3:2 divider on the 270fsb
Tryed 2.5-3-3-7 @ 2.85v but not stable.
I have not tried 3-4-4-8 timings but maybe later.

 

[oldfart]

Looks good so far. Seems your 3200 LL wont go too high. My XMS 3200C2 (not LL) is good for ~ DDR426. Is that about where yours poops out? When you went to 270/3:2 did you go back to 2-6-2-2? What Vcore do you need for 270 FSB?

 

[happybob]

jhites, sorry to ask again, where did you get your intel 2.6c 800 from?

 

[jhites]

Originally posted by: oldfart
Looks good so far. Seems your 3200 LL wont go too high. My XMS 3200C2 (not LL) is good for ~ DDR426. Is that about where yours poops out? When you went to 270/3:2 did you go back to 2-6-2-2? What Vcore do you need for 270 FSB?

Yep the 3200LL pooped out at DDR416.
I did not try anything between 260-270 but was able to get 2-3-2-6
at the 260fsb with 2.65v on the ram. I went back to 2-2-2-6 with the
3:2 divider and set the vdimm back to 2.55v for the 270fsb.
Was able to hold vcore = 1.575v BIOS; 1.554v Actual thru 260fsb.
Had to go to vcore = 1.675v BIOS; 1.65v Actual on the 270fsb.

Been playing with it a 270fsb for a while and will be going to 280fsb
next up. Don't think there is a lot more in the 2.6C now that I started
uping the vcore. I won't run it above 1.70v actual and prefer 1.65v for
safety.

Got some HyerX PC3500 coming Monday and will be seeing if it does
and better than the Corsair.

 

[jhites]

Originally posted by: happybob
jhites, sorry to ask again, where did you get your intel 2.6c 800 from?

MyInfinity.com @ $257 + free shipping.

 

[mechBgon]

Another benchmark that would be interesting is the Ars Technica reference SETI@Home work unit. This isn't a quick test, it would take ~2 hours on a very fast system. If anyone's interested, here's how: http://www.teamlambchop.com/bench/benchfile.htm

You'll need the i386-winnt-cmdline.exe client from halfway down this page. I made a batch file with the line

setiathome-3.08.i386-winnt-cmdline.exe -stop_after_process

and put that in the folder with the client (and the reference work unit, don't forget that you've got to put it in there!) for convenience. For reference, my AthlonXP 1600+ managed this in about 3:15 back when I had it OC'ed to 166MHz FSB on my A7V333-R.

When you start S@H it will come up after a few seconds and prompt you to either use an existing S@H account or begin a new one. Give it an email address and it'll get to work. There is no visible output, ordinarilly. When it's done crunching, the -stop_after_process switch will shut down the client and you can examine the result file for the number of seconds that the analysis took.

S@H is responsive to changes in CPU speed, FSB, memory timings, latency... great CPU/northbridge/memory benchmark. Those of you who are so inclined, hook up with the AnandTech S@H team so we get boosted up the ranks (currently we are #5 in the world, more info in the FAQ I wrote up).

Just to reiterate this point: if you want to do this benchmark, don't forget to put the reference work unit into the folder with the client and the batch file that kicks it off with the -stop_after_process switch. Otherwise the client will happily go download some other work unit, instead of processing the reference work unit.

 

[Duvie]

I am glad to see someone testing a 2.6c, but I wish ppl would actually test real programs instead of synthetic crap like sissoft and 3dmark.....

I want to see mp3 encoding, mpeg2 or mpeg4 testing......

The real important test is ofcourse where the IC7 is going...i would be interested with the better ram if you can test that board at higher fsb...Are you still at 1.55v with 270 or 268fsb??? I am seeing some of these c chips doing 3.6 to 3.8ghz so near 300 could be in yuor range...I think the 2.4c's will likely puch the 300fsb barrier a bit more...

 

[THUGSROOK]

Duvie ~ show us a link to a real encoding benchmark?

out of those benchies he ran id say that pcmark 2002 probably shows the most "overall" realistic results for this comparison.
sandra always exagerates greatly

great job jhites

 

[Duvie]

Well if you mean one that will keep a score for you or time it I can't as most wont....doesn't mean you can't use it!!!!

I ran sissoft yesterday anmd it gave me a score 700 above what it should have been in cpu score scaling it to the 3.06ghz....I ran it 3-4 times and it didn't budge...I ran multimedia and it goes back to correct scaling...I then reran the cpu score and it went back done 700pts to where I expected it to be in the first place, with 3-4 tests showing very consistent number....I call that crap!!!! heck you need to do it on a fresh reboot to get consistent numbers most of the time...

Specviewperf2000+ can be downloaded for free....cinebench as well for some cadd rendering type of software...the prime bench as well as pifast....take a large 100+mb wav file and run that through a host of different mp3 encoders and time it with a stop watch...it can be done but I am not impressed with 5000 in memory bandwidth when I know thay in some of the programs I have tested may be a very small leap or a few seconds increase for a same speed cpu....

As a review I like it for the testing of the fsb on the abit board...for everything else I am not impressed with this and many of the reviews.....

 

[oldfart]

TMPGEnc gives you an encode time.

 

[Duvie]

I know my divx encoding with GKnot does as well but it can be difficult for many to set up and I don't want to get into building a guide for it...The guide at www.doom9.net does a great job but still some have issues....

That divx program lets me encode ac3 files to mp3 with time score as well as gives me time for video encoding in seconds and FPS....

 

[oldfart]

With TMPGEnc, the easiest way would be to just load one of the included templates.