Speaking of CPU's not yet shipping: this is an interesting snippet from JC's:
-------------------------
Yo. What I had hoped might never happen seems to now threaten to occur. Intel's IA-64 project was originally
slated to come out much earlier than it currently is scheduled. Indeed, a snippet from six years ago reported a Intel/Hewlett-Packard P7 Chip" expected to be completed in 1997. A much less vague article from CNET
suggested an initially shipping date of late 1998 for the 64-bit processor. At 500MHz, the chip was expected to
"give PCs running the next version of Windows NT almost unimaginable computing power" . Additionally, according to MicroDesign Resources (the µProcessor Forum guys), "Intel [was] also expected to include hardware
to make sure the processor delivers high performance for older applications which have not been recompiled for the Merced processor". I'm going to file this paragraph under "oops" for future reference.
Just for sake of completion, I'm going to recall for you this 1996 link, which updated the Intel roadmap of the time.
According to MDR, Willamette was slated for 1998H2 and Merced would have come out in 1999H1.
But enough raging here. McKinley. I had lost all faith in Merced fairly early on. The supposed wonderchip simply
had no beginning in sight, and was being delayed every quarter by a quarter. But I saw (and still do see) a better
future for its successor, McKinley. Granted, I may be a little biased here
-- I've known a couple people who've worked on that project and they're certainly smart (and nice) guys
-- but I think that McK will at least be a step in the right direction. But now, something has happened which
seems to have put this second generation IA-64 chip at least into a little bit of doubt.
ISSCC is slated to start up this Sunday. This is a rather important conference in which companies give out pdfs
and presentations on highly technical semiconductor stuff. FYI, they'll be having paper discussions of
microprocessor technology on Tuesday and Wednesday, although there's a bit on future scaling of cpus in the
future on Monday.
According to this article, Intel recently pulled their paper on McKinley from the ISSCC. They were supposed to
more or less give us a full rundown of McK's makeup there, and it's a bummer that we're not going to be able to see the details. Beyond this, some are inferring that this might mean that the release schedule for this new chip
may be pushed out a little. I'm hoping that there's little material effect here. A timely McKinley is something Intel
needs to prove that IA-64 has a reasonable future (though I will admit that the 7 stage instruction pipe does have
me a little concerned).
jc-news.com
-------------------------------------------------
I'd love to hear your slant on things.. tho no doubt you'll find a positive spin out of this -I'm just curious what that might be
:) |
