DirectX 10 + 1900XT

[lhuser]

The thing is: The R600 series with the 80nm technology is because they have still a bit of troubles to make it 65nm. As for the self-looped water cooled 8800GTX, it would eventually evaporate. Hey, why not make it spin and actually have an alternator connected for self-powering?

 

[alvino]

Yeah, of course the water would evaporate. You do know you can still refill the water and coolant, right?

The R600 will most likely be 80nm, but if ATI gets good enough yields from 65nm, it's certainly not out of the question. Look what happened when they gambled with the R520 and the 90nm process.

 

[ownage]

if it is a closed loop, water will not evaporate. The only way for water levels to drop is a leak.

 

[alvino]

No, it depends on the quality of the parts. If you have something shoddy like the ThermalTake Tidewater and it's thin as hell tubing, the water is bound to evaporate (even though it's at a slow rate).

 

[ownage]

No, closed loop = no evaporation.

 

[alvino]

Hmm, I guess X-bit Labs is wrong then. After all, they did review the ThermalTake Tidewater.

 

[ownage]

At least there shouldnt be evaporation. Unless there is an opening somewhere. Hehe, Thermaltake products...no comment about their "watercooling" loops talk about major major corrosion

 

[alvino]

Well what X-bit said was that because the rubber tubing was thin, there might be some evaporation of the coolant over time. Rubber and tygon are different.

 

[lhuser]

The only troubles though is that the water will eventually get hot, even with the fan, so I don't know.

 

[alvino]

Which is why it gets looped into a small radiator cooled by a fan. It's similar to ThermalTake's Tidewater GPU watercooling setup, if you've seen it before.