AMD's ClawHammer is "Cool"

[AGodspeed]

The Inquirer reports that Linear Technology Corporation has released detailed information on their PSU, which is "designed to power AMD's new 64-bit Hammer CPUs. "

Reading through the entire document, I came across some more interesting info:

The LTC3719's 5-bit VID code corresponds to the core voltage range of AMD's Hammer CPUs. Core voltage range is from 0.8V to 1.55V and core current can range from 15A to 40A.

Seems like a pretty low VCore to me. Perhaps the .13-micron SOI process combined with the ClawHammer's integrated heat spreader is having a significant effect.

 

[Utterman]

<< But, the firm also says that the single chip helps thermal management (cuts down heat), eliminates hot spots (cuts down heat), and also removes or helps to remove the need for heat sinks. >>



does this mean you can use passive heatsinks on the hammer?

 

[cjsketchy]

Aww, I thought it was the new AMD slogan designed to combat "Dude, I got a Dell"

 

[zzzz]

Is this ***Confirmed***?

 

[slybender]

<<does this mean you can use passive heatsinks on the hammer? >>

Probably can't use only a passive heatsink because it will generate at least 12watts to 67watts of power within the range by the Linear IC. The heatsink will still need some form of active fan cooling, or heat pipe cooling, or water cooling to remove the heat to avoid overheating and shutdown. To get to the lower range of watt output by the Hammer, you would have to do some serious underclocking by reducing the core voltage to 0.8V, more likely you will be able to underclock to about 1.1V - 1.2V range and still keep the Hammer at a respectable clockspeed.

What would be cool would be the implementation of PowerNow! in the Hammers and a temperature controlled fan/heatsink combo that will turn the fan on if the heatsink reaches say about 50 degrees Celcius otherwise it will be off. So for the most part the CPU will not require the fan to spin, unless the system is being used to its utmost potential cracking RC5 or Folding or playing intensive games.

 

[AGodspeed]

<< Is this ***Confirmed***? >>

The document itself wasn't influenced by The Inquirer in any way. The Inquirer, in this case, is just reporting news, just like they reported Intel and AMD's price cuts a couple days ago.

 

[ElFenix]

i'm getting 62 watts at peak with 1.55V*40A

but thats still a lot of energy to be dissapated from what will likely be a tiny piece of silicon.

 

[AGodspeed]

<< i'm getting 62 watts at peak with 1.55V*40A

but thats still a lot of energy to be dissapated from what will likely be a tiny piece of silicon. >>

Yes but an integrated heat spreader changes all of that.

 

[SuperSix]

<<

<< i'm getting 62 watts at peak with 1.55V*40A

but thats still a lot of energy to be dissapated from what will likely be a tiny piece of silicon. >>

Yes but an integrated heat spreader changes all of that. >>



Really..

So the primary function is to spread the heat? TThe #1 reason there's a "spreader" is to eliminate core damage. There's some small benefit to "spreading" the heat, but it still has to be dissipated somehow.

 

[Mikewarrior2]

An IHS' primary function is for cooling, but it doesn't solve every problem. It does have a great benefit of increasing contact surface and preverenting core damage though.


Mike

 

[SuperSix]

<< An IHS' primary function is for cooling, but it doesn't solve every problem. It does have a great benefit of increasing contact surface and preverenting core damage though. Mike >>



For Intel, the primary function is heat spreading.

For AMD, the primary function is to keep from cracking the cores, and it's good for spreading the heat.

AMD may not say it, but that's why. I've been hearing about an integrated IHS for 2 years now from my AMD rep.