Re: How could a strong Tualatin destroy Itanium, Foster, and McKinley
By offering equal or better performance, without requiring new software, Rambus, or both. Since the Xeon PIII core is currently available with a large cache, there certainly isn't anything to stop Tualatin from being produced the same way - except that it eliminates one of the few excuses for buying one of the new cores.
Re: It isn't intended to scale in clock speed
I know. Which would seem to leave Austin as a money maker (with sub 1.5GHZ Durons) for the next 2 years, even as Intel is paying for new .13 FABs to produce parts to compete with the output of AMD's old and paid off FAB. But if Tualatin did scale, it could provide tough competition for Durons from AMD's Austin FAB, as well as P4, Itanium, Foster, etc.
I see your point, though, that if Intel refused to produce a Tualation with more than 256K cache, it couldn't compete with any large cache server chips from anyone (unless Micron produced a chipset for it with L3 in the chipset)
Regards,
Dan |
