CrystalMark 2004 Competition! Stress your System!

[Bona Fide]

I like the idea of running UT while running Prime95. At least there will be something interesting to watch now and then

 

[rise]

thanks Z . i'm going to try a bit of tweaking on thses 1 gigs again soon. i'm at 41ns i think now.

 

[monster64]

New score, over 32000, but its because I had a very high fsb, yet kept the memory at the same speed using a divider. Zebo, on your pics I saw my memory score was higher than yours, but your other two, the cpu intense benches, were higher because your fsb was higher. Im sorry, I just had to beat you, wouldve been embarrassing if a slower cpu beat me.
http://img235.imageshack.us/my.php?image=myscore29nd.png

 

[Zebo]

:thumbsup:

Originally posted by: monster64
New score, over 32000, but its because I had a very high fsb, yet kept the memory at the same speed using a divider. Zebo, on your pics I saw my memory score was higher than yours, but your other two, the cpu intense benches, were higher because your fsb was higher. Im sorry, I just had to beat you, wouldve been embarrassing if a slower cpu beat me.
http://img235.imageshack.us/my.php?image=myscore29nd.png

Interesting run 1:1 again please and post screen. eg. 250 x 11, same timings.

 

[monster64]

K, here it is, I still got more points than you, however less, than the 305x3. Mustve gotten a bad read last time. However, on avg I'd say the 305x3 is still better, because I got anywhere from 30,900 to 31,700 at 250x11, but doing 305x11 gave me a much higher score general.
http://img15.imageshack.us/my.php?image=myscore33go.png

 

[Zebo]

Thanks monster I don't have my comp ATM but very interesting.. I did tests months back showing HTT means little if anything. What was your LDT set at during both test?

 

[monster64]

Yes, HTT doesnt mean anything, but the cpu interntal clock does, I think. For the 305x9 it was at x3, for the 250x11 it was at x4. So it cant mean anything since it was higher for the 250x11 yet I got a lower score. Also at 305x9 my memory read/write tests showed about 70-100 more mb/s than at 250x11. Cache too. My ALU results didnt really change much, however the FPU test results were significantly higher.

EDIT: Notice how daddy's score was about 500 points higher than the guy below him who had the exact same cpu/speed, and an even higher memory speed, but daddy's internal cpu clock was higher so he got about 500 points more.

 

[rise]

http://www.neoseeker.com/Articles/Hardware/Guides/athlon64oc/2.html

 

[monster64]

I was looking for that, thx. That would explain part of it, but I still got more mb/s with 305x9 than with 250x11. Guess thats because the divider kept the memory at the exact same speed at 305x9 as 250x11 without the divider.

 

[MADMAX23]

Updated monster64 new score of 32009: Test its stability with Prime95 blend test, please...then report your experiences.
"chinXgaX": You are XXXXXXX, ok? The reason is your nick is a censored word...
Thnx guys, I'll keep this thread updated daily till you want/need.
Technonut: Yeah, I know SP2004 is a different GUI for Prime95 but I prefer the original.

 

[monster64]

You want a pic?
I know 250x11 is stable (I tested it when I first got my PC/began ocing)
305x9 is the same speed, so I dont think it really matters.

 

[slag]

still testing

 

[AkumaX]

did a bios upgrade, now scores went up a little: 24716

hopefully new bios will provide higher o/cing

 

[coomar]

tried zebo's prime95 with 3dmark01 looped, seems a pretty good test if you know your videocard is stable, makes sure 3D is stable

 

[Fytron]

32036
Got only 28k at first but I overclocked it from 3.53 to 3.69 ._.

 

[MADMAX23]

Scores up-to-date!!
Added Fytron; Welcome to Anandtech!!
Akumax: new score edited!

 

[monster64]

Fytron youre going down.

 

[munit]

mark: 36059
ALU: 12241
FPU: 14705
Mem:9113

dual opteron 246 at stock 2ghz speed with 2x512mb pc3200 corsair ecc ram at cas latency of 3

well did a bios update and got some better scores

Mark: 38444
ALU: 14445
FPU: 14561
MEM: 9438

 

[stevty2889]

Dothan @ 2.4ghz 24836

ALU: 8634
FPU: 11177
Mem: 5025

 

[Zebo]

So much for weak FPU.. if you had a 730 mem would get way up there too.

 

[EightySix Four]

Dothan at 2ghz

yeah weak FPU doesn't describe it well.

Mark, 21471
ALU, 7466
FPU, 9278
MEM, 4727

 

[Sweeper]

Mark: 25305
ALU: 8662
FPU: 8562
MEM: 8081

Windows XP Pro
P4 2.4C w/ HT @ 3.2GHz 267x12
MSI 865PE Neo2 LS
512MB Kingston HyperX PC3000 @ 427MHz

 

[Velk]

Mark: 55,201
ALU: 19,535
FPU : 22,824
Mem : 12,842

Specs are : X2 4400@ 241x11 (2651), RAM is 2GB 2.5-3-3-7 at a 9/10 divider.

 

[Duvie]

Originally posted by: coomar
yeah i use sp2004, the only problem with that is its fixed to your computer date, if you play with the time (or reset the bios thus resetting the clock) while its running the counter will change

I can see that, but you are running the benchmark app which is not looking for instability (rounding erros, etc)...Run the torture test for a few hours and then see if it doesn't error out... ant test that last like 1-2 minutes is not enough...

 

[coomar]

duvie what are you talking about, sp2004 is just prime95 with a different graphical interface, i like it because I can tell how long its been running since it has a clock feature as well as it records cpu temp every couple minutes so i can scroll and see what temperature it was running at overnight instead of mm5's high/low/average/current