Some serious mark scores?

[Jessman512]

Is it just me or are these scores nothing short of phenominal for my setup?


http://img370.imageshack.us/img370/1...rk055cx.th.jpg

http://img370.imageshack.us/img370/6...rk036gf.th.jpg

http://img370.imageshack.us/img370/3...se011sy.th.jpg

SETUP:
Barton 2800+ @ 2460Mhz (12x205fsb @ 1.925vcore)
|2x512| Dual Channel DDR @ 205 Mhz (410 DDR) 2.5-3-3-11
BBATI X800 Pro @ 587/574 (stock voltage and cooling)
Epox 8RDA3+ w/ latest official bios
120Gb WD ATA100 HD 8Mb
WinXp Home sp2 w/ cat 5.12cp / dx9.0c

[QSDT]

Not bad at all. Have you unlocked your multiplier and if so which method did you use?

[wild_andy_c]

They're okay benches for the hardware indeed! You've used your parts to good potential.

CPU is really a let down though - me and Bradmax57 have topped 30,000 in 3dmark01 with X800's.

[Jessman512]

Yeah cpu, is well old, but for all the parts used i must say, its a responsive system, see if I can drag some more out of it soon.

[Per Hansson]

Really nice... I have a very similar system:

Mobile Barton 2500+@11,5x200=2305Mhz 3x512MB@2,5-3-3-8 X800XL AGP 256MB@400/490 (core/mem, no oc...) Catalyst v5.12 Epox 8RDA3+ BIOS 07/29/2004 with Sil BIOS v4.2.50 1x 74GB Raptor system drive

This system scores 4853 points in 3dmark05... I have not tried overclocking the X800XL yet, it is brand new with a Zalman Fatality cooler on it...

[Jessman512]

Change the trp timing frm 8 to 11 it gives more performance of nforce 2 boards, saving the explaination, can simply verify it w/ sandra.

Card should oc well, in my case i think i got a great x800pro, grabbing a beeter cooler and gonna vmod it soon, see just how much i can suck outta this thing.


3dmarkse01 score is just ok, basically cause cpu is maxxed, but the 03/05 is where the system is shining, for all the parts included, i can still run COD2, Hl2, and even FEAR maxxed settings smoothly.

[madcatmk3]

nice, an AMD 4000+ with a 490Mhz Gtx will get only 8600 in mark 05.

[Per Hansson]

Originally Posted by Jessman512

Change the trp timing frm 8 to 11 it gives more performance of nforce 2 boards, saving the explaination, can simply verify it w/ sandra.

That is not my experience, if I increase the tras from 8 to 11 I lower both my SANDRA memory score and the throughput score memtest86+ reports....

[Jessman512]

Hmm, interesting I was pretty sure the boost from setting to 11 was across the boad (i know it mentally unsatisfying to raise timings). Have you tried lowering your gfx cards timings?

[bradmax57]

Pretty nice scores

[Aedan]

Originally Posted by Per Hansson

That is not my experience, if I increase the tras from 8 to 11 I lower both my SANDRA memory score and the throughput score memtest86+ reports....

11 would be too high - given the figures you've posted, I'd guess 9 or 10.

[Per Hansson]

tras should really follow the speed of the memory, at 166mhz you should be at about 6, at 200 7-8, and so on... Setting tras too low will have a serious data corruption risk, I read this on Lost Circuits years back... They really went into detail on how it worked...

That X800XL I bought sure did not like overclocking, the core tops out at 410 from 400 default, and the memory at 530mhz (490 default) Though I can run ATITools scan for artefacts test at this speed for 12 hours without any corruption... Core is at 63°C max and the board itself (whereever that sensor is..) is at max 49°C (I have the Fatality cooler running at 5v)

Gonna put the ramsinks Zalman included on the card and see if it makes a difference...