TaPaKaH

400-450MHz on QPI gains about 3s with high CPU speeds, don't think that Clarkdale is a competitor for 4GHz category

I had 9:29 on a H55M-UD2H (222x18, qpi4440, mem 888@6-6-6)

That's an amazing achievement, you are in a class of your own, Karl

4000.0MHz is max, 4000.1MHz is not allowed, right?

prolly those submissions were made before they introduced an option for you to tell it's unlocked when submitting - so old entries defaulted to 2 cores

gigabyte employee allowed to participate in gigabyte comp? I already have the board, looking forward to as well

it's a CPU-Z bug that randomly occurs when running your CPU very close to the edge

what's the batch? you forgot about the CPU_OV jumper next to the memory slots?

Turbo=Enabled with 1 core (no HT) gets you x23

it actually isn't, FSBwall 360 air and 435 LN2

http://hwbot.org/community/submission/1018410_ , eh?

just check this guy's other 920 scores and tell me what you see

the only prize is lucky draw?

sry for old result bumping, but what's the batch / pcb code ?

same board as for other benches - Gigabyte H55M-UD2H with highvcore/vtt BIOS from XS

H55 is actually not bad for 32M, I've been pulling 9:2x runs during quick 4GHz pre-testing (222x18, qpi4440, mem888@6-6-6) on air before going to LN2

how much did Andre ask for this score? j/k

golden rule about ASUS 775 boards: if it doesn't die after first LN2 session - it newer will

is this a backup?

situation1 - I've binned 100 cpus, found the best 10, kept one for myself, remaining nine were sold/gifted (as good sample) to my teammembers, they produce scores for the team on their own rig. what then? situation2 - I've binned 100 cpus, found the best one, benched it myself, when done, sold it to my teammember, he benches on his own rig, sells further etc... what then? of hardware sharing prohibition only applies for group benchsessions?

http://hwbot.org/rankings/benchmark/wprime_32m/world_records 1.920 is faster than 1.907 ?

http://hwbot.org/community/submission/855673_ blind has ran 32M at 7714MHz. I had a go with that CPU once - if not for FSB wall in the 335-340 area that would've been by far the best cedarmill out there (i did 8.03ghz 1.82v with that before running into the fsb wall) ... and that FSBwall depends A LOT on the power supply. On Enermax 850 it was 330-335, on HX1000 it's around 320, on OP1000 it was ~340.

in case of all other components being exactly the same the only way to prevent this from happening is to, as said, make SN software-visible (which, personally, I am against since I like to exploit RMA a lot) but in case when other components are different - why not? for example, if I sell/lend (you can't control the moneytransfers) a CPU to a teammember 1000km away, we both make scores on different rigs, we both submit to hwbot - it is ok. if I sell/lend a CPU to my teammember during a groupsession (not that they happen very often in Latvia), we both make scores on different rigs, both submit to bot - this may cause us both to get banned. where is the line? that's why I suggest to allow sharing as long as the benching rigs are not exact identical

will it be for hwboints ?

I think hwbot must not care about sharing of hardware between team members as long as each individual posts results obtained by himself (otherwise - score sharing -> 1 year ban) After all, if a big manufacturer can sponsor* an person - why can't his teammate act as a sponsor* as well ? *sponsoring defined as giving free hardware and expecting something (PR or hwboints) in return