Dan, Re: "Also keep in mind that, last time, Intel management fell for AMD's sandbagging hook, line, and sinker."
Another good reason not to sandbag this time.
But to be sure, I hardly think that 800MHz is showing Hammer's potential. The only way I can see them getting *that* slow is by having a number of analog problems in their clocking scheme, such as in the Phase Lock Loops. I can see the clock frequency doubling once AMD takes care of the major problems at hand, which may all be fixable within a stepping or two of the silicon.
The flip side, however, is that these slow chips gate the validation effort, which should already be going on with both AMD, and their customers. Sure, getting beyond 800MHz will be easy to fix, but the longer this takes, the more the overall schedule will slip.
2.0GHz still seems like a long shot, IMO. I can believe that AMD's simulations suggested that they can reach these speeds, but clearly they will have hurdles to overcome with SOI, as well as their current .13u problems. AMD is under a lot of pressure to reach 2.0GHz so that they can distance themselves far enough away from Northwood. If all they can launch with is 1.6GHz, then the Hammer launch will be a non-event. A 3.0GHz Northwood should have no problem outperforming a 1.6GHz Hammer, even if the IPC is all that it is hyped to be - and more.
We'll have to see how successful AMD is in reaching their megahertz goals. Frankly, it could make or break the Hammer launch.
wbmw |
