Joe, Re: "The economics are just against a niche, low volume products."
But I was under the impression that even niche, low volume products could be significant revenue and profit generators. It's because of this that I think that Intel is continuing on IA-64 development.
Prior to Opteron, I considered it Intel's goal to manage the prospects of their Xeon line to slowly move customers onto IA-64, but now, AMD has a much more effective Xeon competitor, which will force Intel to increase the value proposition of their IA-32 line. At first, I was concerned that this could be a major blow towards IA-64, since as you say, it is part of the "everything else" category.
However, the more I see, and the more I consider, it seems that Itanium 2 may have the performance and scalability to be much more of a compelling product than Xeon, as long as the software migration continues to happen. If this is the case, then I presume that Intel will continue to proliferate IA-64 towards more commodity pricing, as customers begin shrinking away from RISC product lines in search of a lower cost model. There is no reason why IA-64 can't follow x86 in terms of long term cost structure, and still have performance that justifies a new architecture. It simply takes time, though I'm sure you'll agree, time is off the essence here.
That's why I think that IA-64 still has a chance, albeit a smaller one than it had several years ago. Opteron will be a mighty Xeon competitor, but chances are that it won't fit in the "everything else" category that you mention, while it's likely that Itanium 2 will. The real difficulty is when the "everything else" category shrinks, and Intel has already saturated their share in this market. At that point, it will become necessary for IA-64 to migrate towards volume markets, and I agree that this will be a hard sell if software is still in an immature stage of migration.
wbmw |
