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: Windsock who wrote (76362)	4/3/2002 2:54:48 PM
From: Yousef	Respond to of 275872

Windsock, Re: "The failure of AMD to manufacture fast transistors resulted in the need to abandon MHz and adopt QuantiHurtz." AMD is also "abandoning" manufacturing as well ... They're turning this "responsibility" over to UMC. Guess AMD will have to just focus on the "arcitecture thingy" and "new stickers". <ggg> Make It So, Yousef

To: Windsock who wrote (76362)	4/3/2002 2:57:52 PM
From: Yousef	Respond to of 275872

Windsock, Re: "Intel is the technology leader and manufactures the fastest silicon transistors in the semiconductor business." CORRECT ... Looks like AMD is behind by almost a "generation" in process technology (again and again). I'm sure UMC can help them "catch-up". <ggg> Make It So, Yousef

To: Windsock who wrote (76362)	4/3/2002 2:59:23 PM
From: Win Smith	Respond to of 275872

Me and Magee understand that Intel has quite an entertaining army of suckups at their disposal, anyway. History indicates that management will be sure to intervene if said army ever happens to leak any real information, though.

To: Windsock who wrote (76362)	4/3/2002 6:14:21 PM
From: Ali Chen	Respond to of 275872

Windsock, "The failure of AMD to manufacture fast transistors resulted in the need to abandon MHz and adopt QuantiHurtz." You are repeating the mistake made by Yousef. He also claimed that AMD was behind in MHz in the K6 era due to poor "transistor architecture". In fact, the opposite was true, and likely continue to be. The lead of Intel in MHz is due to different partitioning philosophy. Intel decided to break MHz using a crazy 20-stage pipeline, which is not considered as being optimal from scientific point of view. However, Intel resorted to usual bullying. Using their wealth and industry "influence", they re-defined most benchmarks (e.g. BAPCO sm2002) to suit their architecture, by introducing "streaming" tasks that are not on the typical arsenal of majority of computer users. You may want to call it as "vision", but if you think that ripping DVD and CD is what all people are going to do, you may want to hear what movie-producing associations are thinking with this regard. "What you and Magoo don't understand is that Intel will have 40 nanomter transistors with its 0.09 micron process." So far we yet to see any mass-produced results with 130/60 nm technology. I hear that the variation in feature-size control (due to natural thermodynamic fluctuations I guess) lead to excessive leakage of every part, which raises a big question about manufacturability and yields. The problem will only grow with 90/40nm transistors, especially when the transistor count goes to 100,000,000. Therefore, your optimism about "Teratransistors" may be unfounded. - Ali

To: Windsock who wrote (76362)	4/3/2002 8:28:07 PM
From: Bill Jackson	Read Replies (2) \| Respond to of 275872

Windsock, Remember the botched gates? Notched gates, that is? They gave Intel abysmally low yields for just those statistical thermodynamic reasons mentioned by Ali. As you get smaller and smaller those fluctautions loom larger and larger and become the dominant failure mode. Since it is a meander around some mean value you will get trapped in a cleft stick as you get smaller and smaller. It takes only one transistor out of ~100 million to cause a failure(some areas may be more tolerant??). At some point you reach zero yield. Now there is also the thermal effect. At 40nm the devices are less tolerant of thermally induced diffusion of assorted dopant atoms, this leads to lower operational temperatures which leads to lower speeds. Can Intel solve these almost intractable problems? Can AMD? These are some of the questions. Bill