HM, Re: Do you see IPF as being a profitable venture at this point?
That's a loaded question, but I'll answer it intelligently.
Profitability has many definitions, but the important one to you and me might be if Intel ends up with more money if they continue to invest in Itanium vs if they don't. I think this number is a positive value, because I expect the bookkeepers at Intel to scrutinize it very carefully.
It's already a solid choice for the technical compute market and the transaction processing is unbeatable except for the highest end IBM Power machines. HP's share alone, I think, is enough to break even on the R&D.
Of course, then there is the subject of total investment, including marketing, software optimization, chipsets, supporting infrastructure, etc etc. That's a subject that is harder to quantify, but I've heard some very good perspectives on it.
First, it sounds like Intel's software optimizations group, which was started because of IPF, is now used to optimize software for Pentium 4, Pentium M, and even X-Scale, so Intel has been able to leverage the money spent into many other divisions. The return is thus greater than what can be realized by Itanium CPU sales.
Second, IPF marketing is part of the overall enterprise marketing group, which was the source of the "Yes" and "Real Server" campaigns. Intel leveraged the "Yes" campaign with their other divisions, too, including the large corporate market for business clients and mobile workstations. It's not as if Intel had to spend all the money twice to cover Itanium. The IPF portion of these campaigns essentially came for free.
As for other areas, I'm sure that Intel's IPF chipsets have probably lost money, and no one will be able to recover the lost R&D on Merced and all the infrastructure developed since 1997 or so... however, that has all been expensed in the quarterly statements at the time. Intel does not have to pay back anyone for the money spent during that time. Sure, it looks bad from a decision making perspective if they can't recover the money pumped into IPF in the beginning, but from a "this point going forward" perspective, it's water under the bridge.
From a purely CPU sales perspective, I think Intel is making enough money to pay for the CPU R&D, so in one respect, I already consider it profitable. As I explained above, that's only a small part when considering IPF as a whole profitable, but the way I see it, continued investment into software and infrastructure development will only strengthen IPF going forward, and that stands to pay back considerably.
