Hammer Cooling solutions ready

[phynicle]

wow the thermaltake coolers for hammer cpu's are already ready
http://www.thermaltake.com/products/VenusMenu.htm
and the best thing is both solutions contians only 40000 RPM's , now what thats that say about the temp on hammer???

 

[Saem]

Most likely, Thermaltake will have to revise the fan decision. SOI, small die size and lots of metal layers aren't stellar when it comes to helping extract heat from a core.

 

[phynicle]

lots of metal layers u say???
how many is going to be on the hammer packaging?

 

[Gubbi]

Most likely, Thermaltake will have to revise the fan decision. SOI, small die size and lots of metal layers aren't stellar when it comes to helping extract heat from a core.

Huh?

Only thing relevant to heat dissipation is smaller die size because of the smaller contact area between the die and HSF. But since Hammers only have slightly larger power density than AXPs and since they will use heat spreaders this won't be a problem.

Cheers
Gubbi

 

[Saem]

Only thing relevant to heat dissipation is smaller die size because of the smaller contact area between the die and HSF. But since Hammers only have slightly larger power density than AXPs and since they will use heat spreaders this won't be a problem.

Well SOI has a insulating effect. So heat will be made harder to extract.

lots of metal layers u say???
how many is going to be on the hammer packaging?

Depending on how you count them somewhere from 8-10 layers. Current x86 MPUs are 6 metal layers. The extra metal layers will trap some heat.

 

[phynicle]

well doesn't the insulating effect of SOI is countered by it's better conductive properties???that's one of the reason the IBM power processors are abled to clock to the speeds they are now

 

[Tahir2]

I hope Hammer's are going to employ heat spreaders.
Also I am fed up of Delta's screaming.... ack!

Need smarter HSF design (e.g AX7) rather than more brute strength (and hence noise).

 

[Saem]

well doesn't the insulating effect of SOI is countered by it's better conductive properties???that's one of the reason the IBM power processors are abled to clock to the speeds they are now

Yes, but frequency has a quadratic effect on power, while reduction in the dielectric constant will only yield a linear decrease.