Re: chic, welcome aboard. I see you've had a visit from the welcome wagon
twister,
I've enjoyed the kind response you all have given me.
One thing SUNW really has going for them is marketing and sales. Also, they seem to make good strategic moves. My company tends to move slow and sometimes makes bad decisions.
Case in point, we still continue to support INTC's initial IA64 efforts. We are developing an SMP board and who the hell knows what else. Merced/Itanium will go down as the biggest debacle in the history of processors. It's a complete joke. From this article yesterday, it looks like SUNW is making the right move and distancing themselves from the platform. Chalk one up for SUNW.
Posted 05/09/2000 5:35pm by Andrew Thomas
Intel says Sun just isn't trying hard enough
New Intel Europe Merced/Itanium supremo, Olivier Riviere, today reiterated claims that Sun wasn't trying hard enough to get Solaris running on the first Intel 64-bit processors due for launch next month.
Mike Fister, Intel vice president and general manager of the Enterprise Server Group, speaking at a press conference in Tokyo yesterday said:
"Users haven't addressed a request for solutions in combination of the Itanium and Solaris. And demand isn't revving up. The IA-64 work from IA-32 applications that run on Solaris hasn't been going well. Sun's move on the Itanium-based Solaris is slow and I believe that Solaris is becoming an OS dedicated to SPARC."
And today in London, Riviere reaffirmed: "We're not listing Sun alongside people like SGI, Compaq and IBM in our presentations anymore because it simply wouldn't be fair on the others considering the lack of effort they've put in."
Riviere listed Linux, Windows, HP-UX and Acer / Bull / Compaq / IBM / ICL / Motorola / Unisys collaboration Project Monterey as the operating systems of choice for the new silicon.
Itanium is on schedule for launch next month, said Riviere, but he would not be drawn on when volume shipments could be expected. He also admitted that Itanium ? on the roadmap for introduction at 800MHz in a few weeks' time - was still struggling to break the 600MHz barrier in the labs.
An Intel spokesman later stressed that the company was 'on target' to ship at 733 and 800MHz in mid-year. My comment: yeah right, just like Itanium was to be released in the summer of 1997, then 1998, then 1999, well you get the picture...
Riviere said that the entry-level 64-bit chip will run 32-bit applications using the main 64-bit instruction unit and not, as widely expected, with a secondary 32-bit CPU on the same die. This will necessarily involve a significant performance hit meaning the first production 800MHz parts will run legacy apps perhaps no faster than 500MHz ? the 1.4GHz Willamette 32bit chip will be available shortly, posing a difficult buying decision for power-hungry users.
However, on double checking with Intel, a spokesman stated that the Itanium Processor would include silicon dedicated to executing 32 bit instructions.
The more powerful McKinley 64-bit processor is still on track for late 2001 delivery and will feature binary compatibility with Itanium, while providing at least 2x Itanium performance. A key benefit of McKinley will be the inclusion of a large L3 cache on the processor die ? Itanium will rely on discrete cache chips, albeit a whopping 4Mb of them. © |
