How much cooler is the .13 micron process?

[somerandomguy]

Could anyone give me a rough estimate on how much cooler a chip made with the .13 micron process is compared to the same chip made with the .18 micron process?

"Ignorance is bliss, but I tend to get screwed over."

 

[Matisaro]

I think it would be simply the heat difference between the voltages used, if the chip runs with 40% less voltage, it will run 40% cooler, this is a guess however.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[POPEGOLDX]

raystonn said AMD chips will recieve virtually no heat or voltage drop by going to .13

the laws of physics disagree....

 

[Matisaro]

Raystonn works for intel, not amd, he has no secret inside information on amd. I respect him but when he speaks of amd he speaks with the same knowledge as the rest of us, nothing more.

~Matisaro~
"The Cash Left In My Pocket,The BEST Benchmark"
~Tbird1.3@1.5~

 

[somerandomguy]

40%? That's not bad at all.

"Ignorance is bliss, but I tend to get screwed over."

 

[lhgpoobaa]

what i suggest is look at the pentium3 @ 1133 (the one that died a terrible overclocked death) and the tullatin 1133 (at 0.13micron)

see if you can find core voltage and heat output figures for both...

work out some ratio for both...

then you can be sure that approximately the same ratio for heat 0.18 vs 0.13 and core voltage 0.18 vs 0.13 will apply for the next athlon AND the P4.

P.S. raystonn is a nice guy, and has interesting stuff to say... but he works for intell and gets free northwood samples.
that i think makes him just the tinyiest bit biased

Religious wars are 2 groups of people fighting over who has the best imaginary friend.

 

[Phelk]

~ Could anyone give me a rough estimate on how much cooler a chip made with the .13 micron process is compared to the same chip made with the .18 micron process? ~

The PIII 0.18 @ 1.13Ghz draws an average peak of 37.5W
The PIII 0.13 @ 1.13Ghz draws an average peak of 27.9W

This is a 25% reduction. Simple arithmetic would mean that a 0.13 PIII can get to (1.13Ghz / .75) 1.5Ghz

However the Tualatin core would have improved heat dispersion (as does the Palomino core) allowing it to go beyond that (they already have 1.6 on the road map).

This is very much a laymans approach to answering your question but heat disapation is a very complex subject.

My best guess calculation when changing die sizes is to take the max of a core on the larger process then multiply by the ratio of the squares.

So for 0.18 to 0.13 use the multiplier 0.18^2 / 0.13^2 = 1.9

Tualatin (0.18u) 1.13Ghz * 1.9 = 2.1Ghz
Palomino (0.18u) 1.53Ghz * 1.9 = 2.9Ghz
Willamette (0.18u) 2.00Ghz * 1.9 = 3.8Ghz

Sony are working on a 0.07u process which gives a further multiplier of 3.5 leading to something like a 10Ghz Athlon and a 13Ghz Pentium4.

<font color=blue> Smoke me a Chip'er ... I'll be back in the Morgan </font color=blue>

 

[Raystonn]

I never said that. If you are going to quote me, at least do it accurately.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[POPEGOLDX]

if i misquoted you, Then tell the forum what you said about AMD's conversion to .13 ... i will paraphrase , you said " AMD has used all its tricks to make the palamino run cooler and at a lower voltage..so the conversion to .13 wont mean much"

 

[Raystonn]

Once again, I said nothing of the kind. I never once made any mention of voltages and heat with regards to AMD's conversion to 0.13 micron technology. I must ask you to stop making things up about me.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Phelk]

Sorry Pope, Ray's thread-cred outweighs yours. Get off his case.

<font color=blue> Smoke me a Chip'er ... I'll be back in the Morgan </font color=blue>