Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: Ian Anderson who wrote (25868)	7/28/1999 6:11:00 PM
From: Zeev Hed	Read Replies (5) \| Respond to of 93625

Ian, apart of being an "accomplished" farmer, my first career, I also happen to have an M.Sc. in Nuclear Physics (thus "isotopes" <g>) and a D.Sc. in Solid State Physics (thus thermal conductivity and properties of semiconductors). But, by the way, you need neither of these to address the question JD was addressing, the given from the release is that K, the thermal conductivity of isotopically pure Si-28 is 50% higher than that of natural Si, and that is (more or less) a fact. There are many places where high thermal conductivity is beneficial, but it does not seems to me to be cost effective in "run of the mill" circuits, the premium for the raw material will overwhelm, IMHO, any performance benefits. Mind, you there is a theoretical performance benefit which is much greater than improved thermal conductivity in isotopically pure Si-28 and that is the greater charge carrier mobility. That reduces heat dissipation per switching and increases switching speed, which for a given performance is more important than "carrying" the excess heat away. (Thus GaAs of course) There are serious questions in my mind if you can get these benefits when you dope back the the isotopically pure epi, however (intrinsic, or undoped SI, isotopically pure or natural, have lousy conductivity), this because the reason for the higher thermal conductivity (and mobility) is rooted in reduction of phonon scattering and electron/phonon scattering when a substance is as pure (including isotopically pure) as possible. If you dope it back, with boron or whatever, the thermal conductivity (even of Si-28) comes back to within 10% of that of natural Si, as far as I know. If one used an Si-28 wafer (what would that cost? $20/sq inch, minimum?), then you'll have a smaller gradient on the thickness of the wafer (with max temperature at the sites of heat generation, the various junctions, conductors etc), but you will still have to withdraw the same heat generated in the wafer and high thermal conductivity will not help if you cannot remove that heat with an air blower. Thus some companies are actually looking at chilling processors actively (and reduce their operating temperatures which indeed improves performance quite a lot), but that would be the same with both type of silicon. Zeev