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 (47905)	1/30/1999 12:24:00 PM
From: FJB	Read Replies (1) \| Respond to of 1571061

With all this talk over the impact of Merced and that blasted six-month delay (you can tell I'm not happy over the Merced delay), the fact is that Merced is going to such a low volume market compared to Pentium or K6/K7. Merced was initially expected in late 1997, so it will be almost three years late at introduction. While the introduction of the P7 (now referred to as Merced) being co-developed by Inteland Hewlett-Packard Co. is at least a year away, speculation abounds on just what this new architecture will do. Merced is expected to be released in late 1997 and volume production probably will not kick in until 1998. However, industry experts already are furiously debating the impact Merced could have on competing chip vendors, especially the RISC camp. techweb.com Sources briefed by the Santa Clara, Calif., chip giant said the company plans to release Klamath-the next-generation Pentium Pro-in the fourth quarter, followed by Deschutes in the second half of 1997. The chips are named after rivers in Oregon and California. For example, the upcoming P7 chip-due out in early 1998-is called Merced. And a top-secret chip in development is called Tillamook, a river near Intel's sprawling facilities in Hillsboro, Ore. techweb.com

To: Tenchusatsu who wrote (47905)	1/30/1999 1:39:00 PM
From: RDM	Respond to of 1571061

Tenchusatsu and Scumbria Thank you for the excellent comments. I agree with many of your comments. They provoked some minor retorts (that I have put below) as I read them. Tenchusatsu <Alpha's biggest threat is Merced and McKinley, but of course, it isn't here yet.> I agree entirely. I believe, based upon what I have heard, the Merced will be possibly be a disaster for those expecting good performance or a low cost part. However, the McKinley is rumored to be quite a competitor to the Alpha. Alpha of course may be extended in the next few years before the McKinley is ready. Investors want consistant quarter by quarter improvements, but large system developments take a long, long time. <Actually, I don't think extended addressing is that hard to achieve. It was done for Xeon without too much trouble (I think), and it can't be too hard to come out with a K7-derivative that has extended addressing.> Some hacked up addressing is trivial to add. However poorly done, afterthought extensions cause software bugs and do not achieve the maximum performance levels. There were cases of this in going from 8 bit, and 16 bit processors to the next generation. The transition to 64 may be a little different, because 32 bits are a lot and few current applications benefit from 64 bits. <Actually, the Merced will most probably be the fastest x86 processors out there, mainly because Willamette is being released about six months after Merced. > I guess that I would not so confident to put Merced at the "most confident" level. I would estimate that it will possibly be the fastest. Scumbria: <Merced is going nowhere real fast. The chip is huge and will not be manufacturable in quantity. The performance will probably be a great disappointment as well.> This is the sense that I have from what I have heard.