Jaan

Hi! One of my teammate kaidtor have a strange problem- he can't submit 3DMark score to hwbot without compare url. For newer 3DMark it is ok, but with 3DM 01 this is problem. He have a validation error right after uploading verification screenshot, so problem may be there? Same story with any attempt to submit 3DMark score to hwbot, he has a validation error complaining about missing screenshot AFTER uploading screenshot and only possibility to avoid this error is submitting score with FM compare url. Screenshot: Can anyone check/reset user settings? Thanks. User profile: http://forum.hwbot.org/member.php?u=16971

Great score and very nice eff, but please show FPS window next time

Yes, it is random, but not so much, massman is right, I've notice same thing. variations I have is +/- 10FPS max, not 50. And LOD does not help in CH. Recently i help my teammate kaidtor to tweek windows and he has nice results. CH-544.7FPS@5395MHz http://hwbot.org/community/submission/2111259_kaidtor_3dmark2001_se_radeon_hd_5870_115792_marks I was very surprized, how sensitive 3dm01 to mem timings and speed with SB. For example, changing mem speed from 800 to 933 give +30FPS in Nature I have UD4 and now testing MSI GD65 and and just can't reach same perf in 3d01, no matter what. All settings, hardware and opsys are same.

No, it is Gigabyte GTX260OC edition (GV-N26OC-896I) Not SOC. I-ve modded it, and I am also guilty of windows tweeking.

# was 1915. CPU is dead now. Funny thing is, that it was able to run two days in high oc mode, but yesterday I put air cooling on it, set ratio 40X, all voltages was default, run OCCT and left running. At morning I see BSOD on screen, restarted, run OCCT again- BSOD almost instantly. Load defaults in bios, restart, but can't load windows, tried with other HDD- same story- sometimes it boot in win, but not stable at all. So seems, that SB is fragile as its other 32nm brothers..

Another L039B470 from local store- up to 5.3 was quite easy, stopped at 5.4 so far. 1024M stable http://hwbot.org/community/submission/2104831_jaan_wprime_1024m_core_i7_2600k_2min_22sec_281ms

2.1V is nothing fo this chips, it is 130nm chip, so even 2,5V is ok. Windows XP Dodatek?? btw New Beetle from Japan was able to run 1M at such speed with his XP-M http://homepage2.nifty.com/NewBeetle/img155.gif and this is ca. 8y ago.

Try to run some single threaded bench when catching cpuz

Thank you! I'm not very happy with this combo jet- 260MHz is max FSB I get. But mem timings is lowest possible and 2x512MB dimms is not so easy like 2x256. Then monitoring Vnb volage with oscilloscope I've see quite big pulsation @load. Adding bigger LowESR+ceramic+tantalum caps help, but can't eliminate them. Other voltages are ok, no drops. But perfomance is ok, you can compare it @2200MHz: http://www.upload.ee/image/603847/pp_92402_2200.9.JPG

Jeah, this is exactly what I did today- found my old dead DF NF2, found faulty parts and hope get this to work tommorow.

Thanks placid! I've alredy try this cpu with LN2- yes, it don't have CB and can bench at 3+GHz. But i'm not happy with resuts so far, becouse I can't get good eff with current mobo and I don't beat SAE score. And I'm hope, that team Katana (NewBeetle) don't see this challenge- with his golden XP-M 2500+ http://homepage2.nifty.com/NewBeetle/img155.gif

Thanks MaJ0r and Arlin I've try to find my old XP-M 2500+ and modded NF7- they was way better- i've bought this cpu with defect, now its working, but nf7 is far from good- Vcore is overloading@2+ Vcore and 225-230Mz is max FSB No matter of lapped NB and mods. Just trying to get alive my old DFI NF2, found 2 dead mosfets so far, with this DFI I have much better chance to win..

nice fsb My NF7 is not that good, and half dead DFI NF2 can do 270MHz, but it Vcore chain is fall down with over 2V Vcore.. But I'm not done yet

And we still have CPU max speed WR at 8.1MHz in CPU-Z "m" means milli, "M" means mega;)

It is a OT for this thread, but there is another small "bug" to fix- according to hwbot, the fastest CPU is running at 8,1Hz . Why? -Small "m" means milli, big "M" is mega http://hwbot.org/hallOfFame.do?type=result&applicationId=13

I've uploded my result of wprime32M lately- yorkfield@4,8GHz- 8.86sec. http://www.hwbot.org/result.do?resultId=663236 And I have 1.2 points for that. Is it realy right number? This is only wprime result with this cpu. And why my result is not listed normaly: here is wprime32M best results (see page 2): http://www.hwbot.org/hallOfFame.do?type=result&applicationId=14 I can't see my result here..why?

But what if hwbot calculate hwboints not from speed, but percent of oc- maybe that would be more honest for hardcore oc-ers?

just searching for wprime 32M results with E6700 and found again that first 12! result is too fast- quad-core QX6700 result in E6700 section.. link

Now quadcore results are removed, but funny thing is that the best result have smallest points number link

Tyrou posted his wprime 32M results with quadcore in E6700 section: I hope, link works It's impossible to do 32M below 10s at 4,4Ghz dualcore here is verification link- this tells same- done by "Intel® Core2 Quad CPU @ 2.66GHz" http://www.wprime.net/score.cgi?id=3566