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: Tenchusatsu who wrote (91259)	2/3/2000 3:47:00 AM
From: Paul Engel	Read Replies (1) \| Respond to of 1575614

Ten - Motorola's Copper Process Will Surely Help AMD So far, Moto's Copper Process is producing Excellent Results at Moto: "Motorola continues to experience major problems producing 500MHz PowerPC 7400 (aka G4) CPUs, with yields down to as little as one per cent, according to industry sources cited by AppleInsider. " Of course, it will be DIFFERENT for AMD ! Paul {====================================} theregister.co.uk Posted 02/02/2000 12:10pm by Tony Smith Motorola's 500MHz G4 yields very low indeed claim sources Motorola continues to experience major problems producing 500MHz PowerPC 7400 (aka G4) CPUs, with yields down to as little as one per cent, according to industry sources cited by AppleInsider. Said sources claim that the problem lies with the G4's architecture and Motorola's copper fabrication process. Regular readers may recall that it was concerns over just these issues -- and the possible effect on the scheduled introduction of the G4 -- expressed by The Register last June that prompted Motorola to contact us and expressly state that it wasn't having any difficulties with either. Our copper process works just fine, senior semiconductor spin doctor Will Swearingen told us, and we've been using it successfully in memory products for some time now. True, but DRAM chips and microprocessors are very different beasts and operating tolerances for the former are far rougher than they are for the latter. In the end, the G4 did debut within Motorola's broad -- some might say, very broad -- "sometime in Q3" schedule. However, AppleInsider's sources claim that as of early December Motorola had shipped just 10,000 G4 CPUs to Apple, insufficient to allow the Mac maker to re-introduce its 500MHz PowerPC, which debuted at the end of August and was later canned before it was due to ship, thanks to both low volumes and the bug that prevented the chip from running at 500MHz or more. Motorola sources suggest that little has changed, and that yields are as low as two or three 500MHz-rated chips per 300-processor wafer. It's worth noting that the same source suggest G4s that rate at 550MHz and even 600MHz are occasionally produced, but in insufficient volumes to ship at that speed. That at least suggests Motorola has finally quashed the 500MHz bug. Comment from IBM Microelectronics sources suggests tends to confirm there are major problems with the current version of the G4. IBM is, of course, now ramping up G4 production following an agreement to produce the CPU for Apple. IBM always said it would be Q2 2000 before the chip entered volume production, so there seems little hope for 500MHz or faster PowerPC G4s in the immediate future. Motorola is said to be working on the next revision of the G4, which, it's hoped, will solve the problems limiting the production of the current iteration. What's not clear at this stage is whether the next release will be a modified G4 or the G4 Plus, Motorola's attempt to rebuild the G4's architecture to allow it to support clock speeds of 700MHz and up. We suspect the latter, since the current G4 architecture, by its very nature (primarily its short processing pipeline) is never going to get much beyond 600MHz in any case, and the way things are looking right now, it can only just reach 500MHz. Then again, the problem may simply be a hurried attempt on Motorola's part to get the chip out at all, and all that's needed is the update that all chips get some six months after launch when all the errata-no-bugs that were left in order to ship the chip on time are finally cleared out. IBM may well be waiting for that six-month revision, which is why despite agreeing to make G4s for Apple last October, it won't actually do so until around six months after that agreement was made. The timing seems appropriate. When the possibility of a major delay to the G4's debut was first mooted, various sources and commentators independently suggested the chip wouldn't actually ship until May 2000 -- which is pretty much when working 500MHz G4s will ship in volume

To: Tenchusatsu who wrote (91259)	2/3/2000 8:46:00 PM
From: Charles R	Read Replies (1) \| Respond to of 1575614

<I think AMD is stretching themselves a little too thin by trying to cover all market segments with Athlon> I agree with this somewhat - not in the context of Athlon but in the context of K6s. AMD has not shown the aggressiveness to pervade Athlon quickly into the mainstream. But, looks like someone finally woke up and there is no longer talk about K6 going well into 2001. <I really can't see how Spitfire is going to be successful with Via entering the low-end and Intel introducing Timna. > Do you expect CuMine to exist once Timna comes onto the scene? <Finally, their future in desktops, perhaps their most fortified stronghold, is in jeopardy thanks to Willamette. > Man, The first silicon is just in and you feel at the top of the world! Get real. <It seems that now, AMD's success depends on Intel's failure to execute. > Do you base this on understanding of what AMD is upto or is that your hope? <But with Intel's aggressive plans for the near and far future, AMD's only hope now is for another Camino-sized fiasco, either with Itanium or with Willamette.> What is this based on? Clearly you don't think competition is sitting around and fiddling its thumbs? Ever hear of Mustang or Sledgehammer?