TaPaKaH

you should ask the mods to move this score into the "GeForce GTX580 PCI" category

I had multiple accounts banned there after pointing out marketing or product flaws in public ... coincidence? As for registration - last time I visited (a couple of months ago), it was something like $1 to register.

Censorship + registration fee do miracles, Dino

impressive Hypers' clock .. for MSI

inconspicuous advertising

there are different PSC: X-series on air will do 32M at 1200 7-11-(7/8/9)-27 with 1.70-1.85V with best samples capable of 1300 9-12-9. Under LN2 with 1.85-2.00V you can hope to do 1200-1250 with 6-11-(6/7/8)-27, but they say that 8-11-x is faster than 6-11-x so there might be no point. T-series are crap, they barely do 2133 9-11-9 so not worth trying. The A3G-P stuff with white stripes on IC sides is usually worse than A3G-A without stripes.

to be honest, I never tried a Gigabyte for suicide validations, but I have a suspicion that on P5E3-Premium, using proper pot and proper paste, this CPU will surpass my 8308

Irony? I see no iron.

did you start a new website called stock-clock-a-holic.ch? seriously, one card in two modes (ref clocks and average OC on air) is more than sufficient .. unless you bend over too much.

why have five different GTX460 cards when the only difference is stock clock?

You're WAAAY too lucky to hit all those gems first try. You complain about testing nine cards and not finding a good one ... some bin ten times more with the same result

On some cards like 8800 GT and HD3850 memory volume makes "some" score difference running at same clocks, but on many others it doesn't and is sometimes annoying, so I agree, those could be merged. But then again, all depends on hwbot's database policy for separating categories. Typical scenario: Popular card A comes with x Mb memory by default (reference design, say). But one day some vendor decides to release an noob-popular model with 2*x RAM, but to keep the price low, they use the crappier chips. When it's other type (GDDR2 vs GDDR3, say) then I won't argue - it needs a separate category. But when it's same type but with higher latency - is it really necessary? It's like those "special olympics", only with competitors being deliberately crippled by individual makers.

...when you feel like this after not buying any new hardware for two days

http://hwbot.org/submission/1053633 http://hwbot.org/submission/661180 two users from the same team contribute HTPP in the same category ... may not be an isolated case but something wrong with the code in general

there are so many "obstacles" in this score - all 564xB China I had or heard of were amongst the poorest clockers - P35 is not the best chipset for suicide validations, at least on ASUS boards I was getting random walls around 7.6-7.8 even with 8G pretested chips - Gigabyte boards are not popular for CPU-Z validation - Koolance pot with crap paste do you have any non-LN2 scores with this chip?

burning Gigabyte is always nice

this could make for a nice amount of hardware points if you have at least a 6GHz CPU

only 20 chips, you're lucky! what's the batch and PCB of this one?

that's why you should select an ES wife (girlfriend) from an early stepping, so that she can understand your interests

by chance, are the older ones week 0949-0950 and "newer" ones - 1006 ?

oh no, M.Beier can't get any more high-end memory samples from Corsair

I hope your ones last longer than mine ... which died 40mins after I got it

memory clock should be interesting

I haven't seen anyone else in the whole world (except you) to have a kit of 2500 Tridents

starts with S, ends with amsung