Fin, <the K7 core isn't going to be able to keep up with the P4 core EXCEPT in perhaps performance/die area.>
The die size argument has been used in the past to demonstrate the "superiority" of AMD's K6-2 over anything Intel had at the time. This was around late 1998. Intel's Celeron (the one with L2 cache, not the initial cacheless version) had a die size of 140 mm2. In comparison, K6-2 had a die size of 81 mm2. The 'Droids argued that because of this, AMD could produce two K6-2 CPUs for every one Celeron that Intel produced. Therefore, according to their reasoning, AMD could afford to price their products much lower than Intel's equivalent.
Of course, Celeron then proceeded to make life extremely difficult for K6-2 as Intel dropped prices, thereby gaining market share and forcing AMD to lose money on their K6-2 strategy. The 'Droids made a huge fuss over Intel's moves, accusing them of starting the price war (sound familiar?) and dumping Celeron at prices below cost. But none of it mattered, for they failed to see the original point: Having the smaller die size doesn't matter as much as they once thought.
It amazes me to see that the same old die-size argument is being raised again, because indications are that the Pentium 4 is making life very tough for Athlon, despite the fact that the P4 die covers almost twice the area. History is repeating itself.
Tenchusatsu |
