Обсуждение: Intel drops 64-bit Itanium
After failing to make Itanium competitive, Intel is now downplaying 64-bit CPU's. Of course, they didn't think that until Itanium failed. Here is the slashdot story: http://slashdot.org/article.pl?sid=03/02/23/2050237&mode=nested&tid=118 Seems AMD's hammer is going to be the popular 64-bit desktop CPU. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001+ If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania19073
> After failing to make Itanium competitive, Intel is now downplaying > 64-bit CPU's. Of course, they didn't think that until Itanium failed. > Here is the slashdot story: > http://slashdot.org/article.pl?sid=03/02/23/2050237&mode=nested&tid=118 > > Seems AMD's hammer is going to be the popular 64-bit desktop CPU. It's really unsurprising; there was /no/ likelihood of Itanium getting widely deployed on desktops when there would be an absolute dearth of desktop software. Think back: Alpha was presented in /exactly/ the same role, years ago, and the challenges it had vis-a-vis: a) Need for emulation to run legacy software that can't get recompiled;b) Need to deploy varying binaries on the substantiallyvarying platforms;c) It's real costly to be an early adoptor of new hardware, so the hardware is expensivestuff. Certain sorts of "enterprise" software got deployed on Alpha, but you never got the ordinary stuff like MS Office and such, which meant there was no point to anyone pushing "desktop" software to Alpha. And we thereby had the result that Alpha became server-only. Why should it be the slightest bit remarkable that IA-64 is revisiting the very same marketing challenges? It has the very same set of technical challenges. It may well be that by the time it /is/ time to generally deploy IA-64, it will have become the Alpha platform. After all, Compaq sold the architecture to Intel, and Alpha already has a mature set of hardware designs as well as compilers... -- (reverse (concatenate 'string "gro.gultn@" "enworbbc")) http://www3.sympatico.ca/cbbrowne/oses.html "Everything should be built top-down, except the first time." -- Alan Perlis
http://www.theinquirer.net/?article=7966 On Mon, 2003-02-24 at 22:25, cbbrowne@cbbrowne.com wrote: > > After failing to make Itanium competitive, Intel is now downplaying > > 64-bit CPU's. Of course, they didn't think that until Itanium failed. > > Here is the slashdot story: > > http://slashdot.org/article.pl?sid=03/02/23/2050237&mode=nested&tid=118 > > > > Seems AMD's hammer is going to be the popular 64-bit desktop CPU. > > It's really unsurprising; there was /no/ likelihood of Itanium getting > widely deployed on desktops when there would be an absolute dearth of > desktop software. > > Think back: Alpha was presented in /exactly/ the same role, years ago, > and the challenges it had vis-a-vis: > > a) Need for emulation to run legacy software that can't get recompiled; > b) Need to deploy varying binaries on the substantially varying > platforms; > c) It's real costly to be an early adoptor of new hardware, so the > hardware is expensive stuff. > > Certain sorts of "enterprise" software got deployed on Alpha, but you > never got the ordinary stuff like MS Office and such, which meant there > was no point to anyone pushing "desktop" software to Alpha. And we > thereby had the result that Alpha became server-only. > > Why should it be the slightest bit remarkable that IA-64 is revisiting > the very same marketing challenges? > > It has the very same set of technical challenges. > > It may well be that by the time it /is/ time to generally deploy IA-64, > it will have become the Alpha platform. After all, Compaq sold the > architecture to Intel, and Alpha already has a mature set of hardware > designs as well as compilers... > -- > (reverse (concatenate 'string "gro.gultn@" "enworbbc")) > http://www3.sympatico.ca/cbbrowne/oses.html > "Everything should be built top-down, except the first time." > -- Alan Perlis > > ---------------------------(end of broadcast)--------------------------- > TIP 3: if posting/reading through Usenet, please send an appropriate > subscribe-nomail command to majordomo@postgresql.org so that your > message can get through to the mailing list cleanly -- Dave Cramer <dave@fastcrypt.com> Cramer Consulting