CeBIT: Stable 5 GHz Prescott on phase change?!?

[Acanthus]

Originally posted by: Thor86
Pretty much confirms one thing. Air cooling has reached its limit. I'm going to guess 4ghz is not attainable even with the LOUDEST air cooling that is available at your friendly hardware vendor.

Supercooling took P4s to 4ghz when they were topping at 2.8ghz... It always dramatically increases the overclock. When you take a cpu below ambient temps (far below in this case with phase change) you drastically increase your overclock ceiling.

 

[smahoney]

I figure trying to cool a Prescott with H2O over 4GHz would provide a method to heat your inground pool

 

[Jeff7181]

The cooling method doesn't have much to do with the frequency the processor is running at, it's all about the heat generated. They need to find a way to reduce the heat generated and then air cooling will be perfectly fine. Somehow I doubt processors will ship with water cooling setups even in the next 5 years. Water cooling is just not practical, they'll just find ways to make processors run cooler.

 

[Acanthus]

Originally posted by: Jeff7181
The cooling method doesn't have much to do with the frequency the processor is running at, it's all about the heat generated. They need to find a way to reduce the heat generated and then air cooling will be perfectly fine. Somehow I doubt processors will ship with water cooling setups even in the next 5 years. Water cooling is just not practical, they'll just find ways to make processors run cooler.

This isnt exactly true... There is a "clock ceiling" for designs that can only be increased with supercooling.

For example, the 2.4Bs hitting 2.8ghz on air didnt have any heat problems, it was just the limit of that steppings designs. But, with vapochill/prometia they were able to hit 3.7ghz and beyond.

 

[StageLeft]

Originally posted by: Acanthus

Originally posted by: Jeff7181
The cooling method doesn't have much to do with the frequency the processor is running at, it's all about the heat generated. They need to find a way to reduce the heat generated and then air cooling will be perfectly fine. Somehow I doubt processors will ship with water cooling setups even in the next 5 years. Water cooling is just not practical, they'll just find ways to make processors run cooler.

This isnt exactly true... There is a "clock ceiling" for designs that can only be increased with supercooling.

For example, the 2.4Bs hitting 2.8ghz on air didnt have any heat problems, it was just the limit of that steppings designs. But, with vapochill/prometia they were able to hit 3.7ghz and beyond.

Yeah, but they do always advance CPUs. I remember reading about an OC in a guy's freezer, but it's pathetically slow to what we can get with air now.