I'm not so sure about that. The thing has 168M transistors vs. 54M on the P4 regular so it's likely to be twice the die size. Plus, it doesns't have as much economy of scale in its manufacturing as the P4 regular and the Gallatin cores aren't speed binned for as low speeds as the P4's so if they only charge $800, I don't think it will be a big money maker or a money maker at all.
Remember, the vast majority of the costs associated with a new processor are the manufacturing, R&D, and marketing. The actual silicon and materials in a processor like the Gallatin run only a few dollars; the bulk of the manufacturing costs usually consist of the facility costs allocated to each processor.
Intel has already paid off its 0.13 facilities and the Gallatin design, so it doesn't have to buy or design anything significant to produce more of the processor, nor does any part of the 0.13 facility cost have to be allocated to the P4 EE -- thus making economy of scale
irrelevant. Intel simply has to reconfigure a few
idle production lines that are currently not doing anything for the company. The most significant cost of the P4 EE will probably be those associated with marketing. I would be very surprised if the the P4 EE cost Intel more than $50-$60 total per chip (current high-end P4 is somewhere around $20-$30).
For all we know, Intel has been producing and binning the Gallatin cores for some time now; there is no reason at all to believe that Intel just began binning cores last week (although they could have in the past 90 days). Intel has such great yields on 0.13um that it is also quite possible that the vast majority of Gallatin cores can be sold at 3.2GHz, just like most new P4 cores below 3.0GHz generally share the same bin as those at 3.0+. The P4 EE obviously won't be subject to the see rigorous validation testing that is used for the $3600+ Xeon MP.
Facebook
Twitter