Poll: Do you think Intel's 'Yamhill' technology is real?

[Martin]

question says it all.

 

[Rand]

I'm sure it's real and Intel does have some developers/engineers working on it as a "just in case" plan. They'd be foolish not to.
That said, I do not expect it to ever officially be unveiled unless the Hammer attains wide tier 1 support in 4-8way multiprocessing servers AND IA64 continues to be a massive disappointment.

 

[Martin]

bump

 

[AGodspeed]

I bet there are a few Intel engineers working on a so-called Yamhill project. However, I just don't know if Intel is going to ever release it.

Unless Microsoft pushes Intel to release Yamhill and/or Hammer becomes a big hit, I don't think it'll ever see the day of light.

 

[MadRat]

Its real but its not x86-64 compatible, more like IA64 compatible. Plus it will be present in a generation prior to Prescott. Why would they wait a whole year?

 

[AGodspeed]

<< Its real but its not x86-64 compatible, more like IA64 compatible. Plus it will be present in a generation prior to Prescott. Why would they wait a whole year? >>

That's not going to happen. Intel won't be bringing down IA64 to the masses for years, if ever.

In addition, the next core revision for the Pentium 4 is the Prescott (according to Intel), which is slated for a 2nd Half 2003 release. Therefore there is no "generation prior to Prescott."

 

[BlvdKing]

I am hearing that the Microsoft support for AMD's x86-64 is damning for Intel's own version of x86-64 (Yamhill). I bet that if Intel comes out with a 64 bit desktop CPU, it will have to be compatible with AMD's if they want support from MS.

 

[sandorski]

"AMD Compatible", I live for the day to see that on an Intel processor!!

 

[ElFenix]

sanders wanted bill to give support to x86-64 but not to "a competing technology from intel" if i remember correctly. i think that "competing technology" would have to be something like a yamhill.

 

[Rand]

<< sanders wanted bill to give support to x86-64 but not to "a competing technology from intel" if i remember correctly. i think that "competing technology" would have to be something like a yamhill. >>



IA64 could just as easily be the competing technology. If IA64 revives from the disppointment that was the Itanium and starts seeing decent sales that would certainly that will be a huge threat to AMD's hopes of seeing the SledgeHammer in 4-8MP systems.

In my mind Yamhill is at least for now a mute point unless AMD sees a decent amount of market penetration of 64bit X86 software, and tier 1 OEM's supporting the SledgeHammer.

 

[MadRat]

((<< Its real but its not x86-64 compatible, more like IA64 compatible. Plus it will be present in a generation prior to Prescott. Why would they wait a whole year?>>

That's not going to happen. Intel won't be bringing down IA64 to the masses for years, if ever.))

I don't see why not. The native IA64 processors are not designed with single CPU systems in mind. Intel only brought out single CPU Itaniums for development purposes, not likely as a serious product. Pentium 4's equipped with IA64-compatibility are absolutely no threat to their high-end product's market.

 

[BlvdKing]

The IA64 would suck for a desktop CPU period. It has to emulate x86 code in software and takes a huge performance penalty for it.

 

[MadRat]

Would you be specific why IA64 would suck for a desktop? I think you are confusing the instruction set with the implementation of the design.

 

[Martin]

<< Would you be specific why IA64 would suck for a desktop? I think you are confusing the instruction set with the implementation of the design. >>



what about all the software that would have to be rewritten?

 

[MadRat]

Why would the old software be rewritten? You can run older software on Itanium as it stands now. The architecture of the Itanium does not favour code designed for processors with OOE in mind.

 

[Martin]

<< Why would the old software be rewritten? You can run older software on Itanium as it stands now. The architecture of the Itanium does not favour code designed for processors with OOE in mind. >>




From what I've heard, 32bit execution is painfully slow.

 

[MadRat]

<<From what I've heard, 32bit execution is painfully slow.>>

It has nothing to do with IA64 instruction sets. It has everything to do with the physical design of Itanium. You're confusing the two.