Massman

I vote 'YES' on giving this chip to Vince so he can try 01.

Bios V1.0B2x Bios V1.0B29 Bios V1.0B31 Bios V1.0B32 Bios V1.0B34 (right-click and save) Don't touch anything before B30. These were very early BIOS releases and contained a lot of bugs. First BIOS that gave me decent results was the B31 (read: DDR3-2000+ with B2B=0). B32 and B34 fix the other subtiming settings in the bios, but you can also check B31 and do the subtimings in CPU-Tweaker I guess. If you test older bioses for better ram clocking, please give some feedback if there's a dramatic improvement so the guys at MSI can look back at a previous bios and see what they changed.

Hm. Seems like the best BIOS for me was still the B32. I'll upload it here and let me know if it's any better/worse.

I forgot to refresh the memory tab after increasing the BCLK through the dashboard ... it's not fake.

Why NB on LN2?

Everyone = 2 None of them say it's "normal".

Hm, just replicated the issue ... very strange! We'll look into this.

Can you make a screenshot of the message? The mainboard is correctly configured in the database IIRC, so it should be ok.

My best guess is that it's a bios that has been tested to co-op perfectly with the Kingston memory that was used. Bios options (eg: voltages) are the same. In my experience, the voltage options are good enough for Gulftown. Make sure you use the OC Switches at the bottom of the board to unlock higher voltages (like 1.7V+ vcore)

What is the full spec of the DRAM? I'll try to get someone to test it in HQ to see if there's some bios issue. Just FYI, here's the V1.21 MOA bios using in ... ehr, MOA EU. Bios V1.21 MOA

30-6-10 23:59:00 CEST The timer on the competition page is always correct

V1.2 official (http://download1.msi.com/files/downloads/bos_exe/7666v12.zip) I'll check for a new bios today. Pretty sure there are new ones out already ... we played with 1.21 at MOA if I'm not mistaken. Beaux: sounds like SPD/bios problem. Have you tried to flash to an older bios and see if there's anything different?

What was the problem?

Haha ... you need to hide the CPU pot in a book?

I think quite a lot know about this. The advantage is that this method is quite elaborate, so the time will be off by quite a bit ... but there's a loophole. We're currently working on an app that would allow more secure submissions ...

The discussion on giving people the points they deserve is a very difficult one since it's more than just "I did effort - I get points". Technically, you should reward people who put more effort into the overclocking achievement more than those who only quickly benched and got a high score. The problem is that it's not easy to make a functional point algoritm based on this subjective feeling of effort ... it's not easy. The hardware points are already an attempt to give more points to those with more low-end hardware. It allows people to bench ancient hardware and still get some points + allows people with hardware that has been tested A LOT (eg: 920) to get some points on air and water cooling. It's already quite elaborate. Why and how should we separate extreme editions?

Actually yeah. Somewhere in the future, we might to try and come up with a point ranking based on platform/technology/architecture rather than CPU class. But ... that's for the future, the plans are not put in action right now.

I wasn't trying to be rude, I'm trying to express my disbelief. I just can't understand how you would come to the conclusion that it's okay .

Why? ... how? Why ... why do you think it's OK to submit in a different category just to get more points? To be 100% clear: no, you have to submit in the right category. That's why we HAVE the categories ...

And then just beat all those scores ... yeah

Knoplerbruce has 918,1 on 1337,4 max points

They will come with a new set of achievements (this or next week I guess)

Hey, Knut ... do you know something about this socket?

Just checking the S939 point ranking ... hwpoints with CPU ...

http://hwbot.org/competition/ocn_gpu_challenge