Replies to post #28733 on Advanced Micro Devices Inc (AMD)

[CombJelly]

"The programs that are losers or just break even with 64 bits will not be represented in anywhere close to realistic proportions."

True. But trumpeting the winners and ignoring the losers is what marketing is all about. But given the fact there probably aren't two people at AMD who can spell "marketing" the same way, much less spell it correctly, this is but an idle observation...

Now at some point, everything is going to have to have "64 bit" on the label regardless of the performance hit. But that isn't going to be for a long while.

How about 64 bit edlin? It's got to still exist out there somewhere. Or QBasic?

[fastpathguru]

"The programs that are losers or just break even
with 64 bits will not be represented in anywhere close to
realistic proportions."

Those that get the 30% boost won't be complaining.

Is this how you're going to spin the 32-bit compatibility feature as a fault? That the range of applications is so huge, and many won't explicitly benefit from 64-bit features, that the average boost granted by long mode will be low?

I'd bet that the doubled register file, low memory latency, and glueless multiprocessing will boost the average application's performance just fine.

fpg