poparamiro

Yes, test them again with kaby.

No, from what i can see in my tests, it is part related to temperature, part related to CPU (one 4890 is ok on one CPU, bugged on other), and who knows what. I have cards with bugg on everything, cards bugged only on some HW used, cards bugged at some level of temp. HD 4850x2 whas bugged on water at 10c and became ok at -30c ==))) S`hit cards from ATi, no surprise here.

Thx, Hehehe, always nice to see you bench. Unfortunately it is hard to find good 4850 in 3dmark 2003, the bugg behavior is terrible. I have two good 4890 and one was 2003 gt2/gt3 bugged. The good one bugged too when i put it to chiller ==))) All the best from me Ramiro

Thx, one card half GT2/GT3 bugged, shit result

Great work, Nachtfalke & I.NFRAR.ED

full scrot ==))))

Agree with Rauf, stop "sucking Intel's cock".... poor Intel ==))))

Great result with that card

Nice work on that vga

Thanks ikki-san

==)), enerveaza-i tu ==))) great chiller man

great cascade demonstration

Heheh... nice man, and crazy collection What ASIC quality on 480?

very good run

awesome, nice pictures also

A little slow

Great one

A lot more resistant VRM that i think Nice result

vGPU only, for vMEM must change inductors.

Gelid die and IHS, no lapping

It is very powerfull, on my gtx 470 session i have -16oC load temp. (poparamiro`s Unigine Heaven - Xtreme Preset score: 1983.52 DX11 Marks with a GeForce GTX 470)

Huge freq for cascade

So, can be done? Single stage is separate cooling category to cascade, waterchiller must be in it's own category, not display as TEC.

I think TEC / Peltier / Water (Chilled/Cold) cooling category must be renamed to waterchiller because very few use tec or peltier. I bench on waterchiller and appear that i use TEC, wich is not accurate. http://forum.hwbot.org/attachment.php?attachmentid=5022&stc=1&d=1480325227 Or better, make waterchiller separate cooling category (to display at cooling method of course). Thanks Ramiro

Great benching