Re: I think what you mean to say is that IPF is the first high end processor to target the server and workstation market with a entirely merchant chip sales business mode - just like x86. In contrast the high end RISCs are basically the house processors of their respective owners and thus employ a vertically integrated platform business model.
What makes X86 servers so appealing is their complete compatibility with X86 desktops and their moderate costs.
Itaniun has neither of those factors going for it. Opteron has both of those factors going for it.
But Itanium has the Intel name, Intel marketing, and Intel's deep, deep, pockets behind it.
I think Intel has been scared by Opteron, and you will see Intel racing at high speed away from their "Itanium is enterprise only for the next few years" strategy for 64-bit computing. Soon the statement "64-bits won't be needed on the desktop until late in the decade" will be a "that statement is no longer operative" statement.
Trouble is, they've spent so much time focusing on "enterprise" platforms with "enterprise" costs intentionally restricted to running only "enterprise" software that it's going to be a scramble for them, Intel name and deep pockets, or not.
And it's tough when you have a competitor with 1/2 an order of magnitude lower costs than yours. A 4 way opteron platform consists of 4 193mm2 CPUs, an I/O controller and a board. A 4 way Itanium III platform consists of 4 380mm2 CPUs a chipset that's almost as complex as the CPUs, I/O controller and a board. Doubling the size of a chip generally quadruples its cost. And the Itanium chipsets (so far, at least) have been complex and expensive.
PS - here's the link: Otellini: Not any time soon. We use that in workstations; there are a number of server-type applications that take advantage of the memory addressability. But there are very few desktop client applications that take advantage of even the full 32 bits today. Even the Pentium 4 has a 40-bit architecture that very few software developers use. Why? Because you don't have the need for memory addressability, and memory subsystems to populate it are terribly expensive.
If you plot the memory requirements of typical applications in terms of their growth, and plot that against the cost of memory subsystems coming down over time, you don't get a reasonable intersect point until very late this decade. http://www.pcworld.com/news/article/0,aid,107238,00.asp
It's important to understand that reserved addresses for hardware and some OS functions mean that a 32-bit processors can efficiently access 2 to 3 gigabytes of RAM - that's it, a 64-bit CPU is required to take full advantage of 4 gigabytes of RAM. What constitutes a "terribly expensive memory subsystem?" Well, that's a judgment call, but 4 gigabytes of RAM, in the form of 4 1GB DDR DIMMs, including shipping and handling, for all 4 DIMMS, is $492 from several vendors on pricewatch.
A year from now, it will almost certainly be less.
Market Data 
Markets 
AI
