<If that's the case, shouldn't you be more worried about Alpha being trounced by Foster, the server version of
Willamette?>
No... two different leagues. Since we are both selling
futures here, the 21264 comes in different flavors by
then and 21364 shouldn't be far behind. Maybe timeframes
are off , but an Alpha OEM writes:
"Keep in mind that 1.5 GHz copper-SOI EV68 21264 to be shipped in systems in Q3 2000 will
reach around 90 SPECint95 and 150 SPECfp95 when combined with 8 MB 1 GHz DDR-SRAM L3 cache..."
That is a server part. Foster might be decent and will
indeed ship in larger numbers, etc. etc... but performance
difference will be quite evident.
Itanium is quite an interesting study... what cracks me up
totally is Linley Gwennap gets away with Itanium suck-up
stuff. Remember these lift-quotes:
linleygroup.com
"We have increased our projections of the chip's SPEC_base
performance to 50 int and 80 fp. We expect it to achieve 45,000
tpmC on the TPC-C benchmark in a four-processor system based
on Intel's 460GX chip set and Lion motherboard. These scores
should give Itanium performance leadership when it is released, but
its performance could be surpassed by Compaq's Alpha
processors within a matter of months."
Linley isn't even interested in what type or speed
Alpha will be shipping when Itanium debuts. I bet his
credibility still holds up, but I have a copy of what
he has written ... we will see how well it looks come
October. By the way, IBM is sampling very fast copper
21264 right now.
"When Merced is in volume, it will outperform non-Intel systems. Merced will outperform Alpha systems."
-- John Miner , Intel Enterprise head
Pick your benchmarks and Alphas VERY carefully... whatever
you do , don't pick Alphas that are new in the same
timeframe as Merced... pick old Alphas and you have a
fighting chance.
Now , putting the shoe on the other foot... suppose
1.2 GHz Athlons with on-chip L2 do better than Itanium,
who buys Itanium and why?
Itanium -- All dressed up and no place to go.
Rob |
