TaPaKaH

looks like a competition of true overclocking skill

0802, as far as I remember...

may be I have missed something really important, but my Gigashyte P67A-UD4 won't flash any of the BIOSes given in the first post - error message is "Invalid BIOS image"

The reason why that record stood for so long is because noone cared about it as it requires a very specific benchrig (hot cpu and cold nb). I actually tried my board with dryice on chipset, anything under 0c and it would not get further than POST with any HTT

Monstru, what "tweak" are you talking about? The inverse HTT/temp scaling was known since the dawn of time, how did you expect the 3000+/3200+/3700+ etc records were broken? Actually, a better idea would be to go with a Winchester chip rather than a Venice, back in 2005/2006 those CPUs were known to have less HTT/cold issues than Venice so higher HTT should be possible as well. (personally, I couldn't find one on time so had to use SanDiego for my submission)

what's the batch/stepping of this 3200+ venice by the way?

you didn't have TCCD ?

Monstru, I know that HTT/temp is inverse, hence I did not even think about CPU on subzero ... IIRC, this O152 does around 380-400 at -60, the molten O144 did over 400, so it'd have been a lot better on air as well

I think everyone is shitting bricks ATM

well, yes ... I had a much better Opteron 144 (did 400+ with subzero, so supposedly 500++ air) but it has molten during yesterdays pretesting

okay, white flag very nice way to end up the country cup (AMD HTT WR)

DFI was not sold in Bulgaria so that you all are using EPOX ?

is that all you've got?

at first I thought those are DDR2 ... and also, can RE really set CL4? I always thought 5 was the min with DDR3

did cold on NB help?

what is the batch/PCB code of this CPU ?

that all on air/stock cooler? I so wish I could participate in this, but seems that none of my 939 cpus are in working condition (2xvenice, 2xopteron)

is this a reference card ?

I wouldn't expect to see CPU to hit over 6.4 on cascade / NB scaling from going dice->LN2 you need to turn the cascade/LN2 the other way around and then give it a go

what's the problem with the hard-voltmod ?

VERY nice board sample you got there

so now we'll have 2M/4M/8M/16M for hardware pts as well ?

I see nothing wrong as long as you don't submit multiple scores from same CPU/session into one category (benchmark) Extremely impressive job there, not many E8600 can hit 6.4 with cascade temps.

highest HTT I know of on AMD is 564MHz by MAZ from Japan

2:3 (720MHz) being misreported as 3:5 (800MHz), which is common for P45/X48 using 400 strap