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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Time Traveler who wrote (31379)	4/8/1998 2:53:00 AM
From: wilee	Read Replies (1) \| Respond to of 1573004

John, I'm new to AMD thread but not the industry. have seen a few of your most informative posts and would thank you by sharing my 2cents. regarding the yields, using the saem assumption and data source: From the AMD chips roadmap: Chip Prcs DieSize #Transistors Frequency Schedl K6 0.35 162mm2 8.8M 233 1H97 K6MMX 0.25 68 8.8 266 2H97 K6 3D 0.25 81 9.3 300-350 1h98 k6 3D+ 0.25 135 21.3 350-400 2h98 My variables: W : usable wafer area, assuming same wafer size for both 0.35 and 0.25 S3: 0.35 die size; s2 0.25 die size Y3: 0.35 yield; Y2: 0.25 yield All other assumption saem to yours: then: # chips 0.35 : #chips 0.25 = 9:1 i.e.: (W/S3) * Y3 * sum(from 13 to 25) : (W/s2) * Y2 * sum(from 1 to 13) = 9 or: (Y3/S3)38 = 9 (Y2/S2)*14; or (Y2/Y3) = (s2/s3) / 3.32 I'd be conservative in my yield estimation and assume all 0.25 chips are the smaller k6mmx: s2/s3 = 68/162 = 0.42 So I get Y2/Y3 = 0.42/3.32 = 0.13, i.e. roughly the yield of 0.25um is of one-eighth that of 0.35. If all were the larger K6 3D, then the ratio gets to 1/6. Assumeing a 80% yeild of 0.35, 0.25 yield stands at 12-10%. That is not at all impressive but at least a good starting point! I think the news is a good one and hope AMD stock goes up. P.S. It would be interesting to see how the yield goes at k6 3D+ with its much larger die size -- don;t you think that it takes too many extra transistors (12million !, more than doubled of 3D!) to go from 3D to 3D+? Could you or any one else on this thread shine some more lines on the architectual/performace difference between 3D and 3D+?

To: Time Traveler who wrote (31379)	4/8/1998 3:15:00 AM
From: Petz	Respond to of 1573004

JYW, you are clueless. Its based on this 'fact' that you use in your analysis: week 13 = 13 parts of 0.25um and 13 parts of 0.25um (claimed by Jerry --- 50-50). Wafer STARTS are 50-50 now, THEY WERE NOT 50-50 in mid January when the late March CPU's were being started. Also, did you notice that Jerry said that MORE THAN 2/3 of the 0.25 um K6's shipped this quarter were from FAB 25 -- yet these didn't exist until March 1 at the earliest! I hope you had fun with your useless arithmetic. Petz