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 : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Bill Jackson who wrote (51946)	8/21/2001 8:32:50 PM
From: milo_morai	Read Replies (1) \| Respond to of 275872

<font color=blue>nForce MB pics m3dzone.com

To: Bill Jackson who wrote (51946)	8/21/2001 10:26:12 PM
From: heatsinker2	Read Replies (2) \| Respond to of 275872

Bill- Well, that bad layout means more clusters and thus hotspots That was Ali Chen's explanation, not mine. I don't know if this is correct. That means that Intel has more to lose with SOI as the extra oxide layer hinders heat transfer. Might be true... This also means that Intel has more to gain from Si28 epitaxial layers. If you form an insulating layer and then deposit an epi layer on it how do you get the epi layer to have a perfect ccrystal structure?, etc. Who knows? I am NOT an expert on this stuff. Sorry man...

To: Bill Jackson who wrote (51946)	8/21/2001 10:58:10 PM
From: combjelly	Read Replies (1) \| Respond to of 275872

"If you form an insulating layer and then deposit an epi layer on it how do you get the epi layer to have a perfect ccrystal structure?" If you are talking about SOI, then the most common process forms the SOI layer under the surface of the silicon. The silicon is later annealed to restore the crystal structure as much as possible. See chips.ibm.com Not sure about diamond (is this done?), but the crystal structure for sapphire is very similar to the of silicon. It is not perfect, so you get some lattice defects when you grow the epitaxial layer, but it can be done.

To: Bill Jackson who wrote (51946)	8/22/2001 12:00:33 AM
From: pgerassi	Read Replies (1) \| Respond to of 275872

Dear Bill: The most common method of making SOI wafers (at least to last year) was to take a cylinder of silicon crystal, split off a layer of a predetermined thickness, implant oxygen by some process to some thickness around that wafer, split oxidized wafer to a thickness of the oxide layer plus some small predefined thickness of the silicon crystal underneath, take the resulting thin piece and bond it to the other thin piece joining both oxidized layers together. You are then left with a wafer with pure silicon crystal on top, a SiO2 layer twice as thick as the implanted layer and the rest of the thickness being more pure silicon surrounded at the edge by a single thickness of SiO2. Using this process (and there are many others), the Si layer is as defect free as the original wafer, as thick as desired and the oxide layer is also as thick as desired. Since the oxide side is placed against a heatsink in the typical flip chip, I believe that the reason for bonding to the other oxidized half is to allow ease of handling and minimize breakage. If a way could be made where that bonding is not done, then two SOI wafers would result from each oxidized wafer with extremely good heat transfer characteristics (pure silicon crystal has a lower thermal conductivity than either aluminum or copper). Pete