McKinley chipset available before McKinley
Numerous problems with this one... I'll start with the big ones and then move down...
The word on the street is that Intel will not be able to deliver the McKinley chips until sometime in the third quarter of 2002, which is considerably later than the late 2001 debut that Intel had been shooting for.
I always love it when an analyst uses the phrase "word on the street." Translation is, "I have no idea what I'm talking about so I'm listening to one guy who sounds like he or she does." Intel has publicly stated several times that, like the Merced, McKinley will have a "pilot release" starting in Q4 of this year and will be officially released (again like the Merced), middle of next year.
The author also contradicts himself:
Intel actually demonstrated the Pluto chipset running in McKinley-based servers at the Intel Developer Forum in late February
How could Intel demonstrate chips back in February that aren't going to appear until this time next year? Doesn't make any sense. Even if there were speed path issues, they DON'T take 18 months to fix!
It is unclear if the basic design of the Pluto chipset is a two-way motherboard, like the HyperTransport clustering developed by Advanced Micro Devices. With HyperTransport, one, two, three or four two-way AMD server boards can be glued together to create two-way to eight-way servers.
Golly, when did AMD release that???? Answer, they haven't...
I think this analysts shorts are showing... |
