Replies to post #21850 on Advanced Micro Devices Inc (AMD)

[sgolds]

kpf, yes, I agree that IPF is a great example of a technology that took just a little longer than promised (by about 5 years).

Now, as a software person, I will stick my neck out and try to justify why it would only take until early or mid 2005 to market a Dothan successor with AMD64 inside:

1. Who said they didn't start yet? Intel has known about their 90nm P4 problems since early in 2003.

2. This is known technology, unlike IPF which had just too many unknowns (and too much handwaving about how the compilers would fix everything).

Integrating dual core & memory controller: The memory controller is not a difficult thing at all, in fact it is pretty much a bolt-on item. Dual core seems to be harder than it first appears because AMD has been threatening to do it ever since the aborted Mustang chip. Maybe it is a die size & manufacturing issue, in which case technology is catching up to it.

So I submit that if Intel has been working on it then an early 2005 release of a Dothan successor with AMD64 and integrated memory controller is doable. I agree that dual core looks like it will take longer, although I really don't understand why.