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: greg nus who wrote (27867)	1/14/1998 2:09:00 AM
From: Yousef	Read Replies (3) \| Respond to of 1572208

Greg, Re: "SDC certified the process for k-6 and transfered the process technology to Austin ..." This is only part of the problem ... I posted this some time ago about AMD's process development model. Sounds like it is time to revisit ... Message 2121363 "I feel very sorry for the people involved with Fab25 in Austin. I have personally been in a similar situation of ramping volume in a process that is not manufacturable (low yields). This is NOT fun. But actually where I feel that AMD has "messed up" is in their Process Development Model. Let me explain, AMD has a development model where IC processes are developed in the Bay Area (Santa Clara) and then transferred to a volume manufacturing site. This model has many problems: 1) What baseline process is running in Santa Clara (SC) ?? Are they running the AMD .35um process at a reasonable volume (as a baseline in Santa Clara)?? If not then the defect density will be high as they do their development work on the .25um process. 2) Is the equipment set identical between SC and Austin ?? If not then new problems will arise as the process comes up in Austin and SC will have a hard time "debuggging" the problem. Sharing of "best practices" and solutions is harder. 3) This is a very expensive way of doing process development !! All the baseline material is typically NOT sold to customers, so the tendancy is to keep the SC volume relatively low and the process can NOT be fully characterized without volume. Also, the Bay Area is about the most expensive place in the country to pay for technical people, real estate for buildings, permitting for expansions is difficult ... Yousef's Recommendation to AMD: Move your process development efforts from SC to Austin. Develop your future processes where you will do your first manufacturing. This allow you to leverage the existing .35um process (or .5um process ...) to establish a low defect density baseline from which to make changes as you go to the next generation. All this baseline material is by definition, production material that is sold to customers. Austin is a much (much) lower cost operating environment as compared to the Bay Area. Then copy the equipment and process exactly as you bring up other mfg sites (like Dresden). BTW, this is essentially the Intel process development model with just a change to the location of cities." Make It So, Yousef