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: dougSF30 who wrote (197196)	5/18/2006 1:20:12 AM
From: Hans de Vries	Read Replies (1) \| Respond to of 275872

Re:90nm Opteron was 114mm^2 after the shrink. Perhaps they then made further changes OK. It's not so well known but the process node beyond 130nm targeted originally by AMD was actually the 100nm node as could be found in various papers from those days. (The HiP8 process together with Motorola).So they were following the usual decimal-exponential scaling pattern: 1 - 0.7 - 0.5 - 0.35 - 0.25 - 0.18 - 0.13 - 0.10 - 0.07 - 0.05 - 0.035 freescale.com Around that time the International roadmap changed to the binary-exponential scaling pattern 130 nm - 90 nm - 65 nm - 45 nm The usual sizes given are 193 mm2 versus 114 mm2 where the latter stems from the 2003 prototypes. Now you can check that: 193 x (100/130)^2 = 114.2 mm2 The 114 mm2 Athlons are rather 100nm ones marketed as 90nm versions. The size has since then gone done to 106 mm2. Now the 65nm node is done entirely in cooperation with IBM so it's really hard to say what the exact scaling will be. Actually 90->65 is a bit less than sqrt(2) so the scaling is about 0.52 Another interesting scaling example is Intels 65 nm process which they could advance well beyond the International Roadmap because, amongs other reasons, they didn't scale the critical Metal 1 layer. This layer, which has the smallest wire thickness, was only scaled 5% compared to the 90nm process. ftp://download.intel.com/technology/silicon/Yan_Borodovsky_SPIE_2006.pdf sheets 22,24,25,26,27 Altough it seems they can nevertheless increase the density by a factor of two for most circuits anyway by clever routing. For AMD's Quad Core it remains to be seen what the actual size will be but it does look like they can get it pretty small. Regards, Hans