seen these "paper" arguments before, wait for proof
fpnc said:
this means when a 64-bit application is run on a 64-bit version of Windows the processor uses 64-bit addresses and has full access to 64-bit hardware registers.
And Windows will have 64-bit graphics and GUI libraries - no "command line" restriction like OSX 10.4 ....
_____________________________
In some ways, these "Intel dual-core is junk" stories are reminiscent of other techno-hype myths from recent years
Claim: G3 will rule - RISC is king, CISC is dead
Truth: After a strong early start by RISC, CISC won the race
Claim: G4 AltiVec is a GFLOP supercomputer
Truth: P4 steadily increased its lead, except for a few sweetheart AltiVec apps (like BLAST)
Claim: G5 has super bus, great memory, will kill the Pentium
Truth: G5 was matched pretty evenly with the P4 (except on really AltiVec-friendly code), but has failed to keep up and is now lagging
Claim: (The initial) dual-core Intel architecture is junk, AMD/970MP much better
Truth: We'll have to see real systems from all three to understand if the claim is true, but it's far too early to claim victory.
The AMD approach looks better on paper, but so do the arguments for RISC vs CISC and for the 970's memory bus. Real world performance will vary by application - depending on how much synchronization is needed between threads on the two processor cores.
Note that Intel will be pricing dual-cores very similar to single cores, so even a 50% boost will give better price-performance.
And, by the way, if the AMD architecture is great - doesn't that mean that Windows will have a great chip? I don't understand Mac people rooting for AMD, unless their dislike for Intel is stronger than their common sense!