Intel Officially Sinks the Itanic, Future of IA-64 Architecture Uncertain

[btarunr]

Intel has unceremoniously, through a product change notification (PCN), discontinued the Itanium family of microprocessors. The Itanium 9700 "Kittson," which was released in 2017, is the final generation of Itanium, and its sales to new customers have stopped according to the PCN. The series has been marked "end of life" (EOL). Existing customers of Itanium who already have their IT infrastructure built around Itanium 9700 series, have an opportunity to determine their remaining demand of these processors, and place their "Last Product Discontinuance" order with Intel. The final LPD shipments would go out mid-2021.

With this move, Intel has cast uncertainty over the future of the IA-64 microarchitecture. IA-64 was originally conceived by Intel to replace 32-bit x86 at the turn of the century, as an industry-standard 64-bit processor architecture. AMD laid the foundation for its rival standard AMD64, which could go on to become x86-64. AMD64 won the battle for popularity over IA-64, as it maintained complete backwards-compatibility with x86, and could seamlessly run 32-bit software, saving enterprises and clients billions in transition costs. Intel cross-licensed it as EM64T (extended memory 64-bit technology), before standardizing the name x86-64. Itanium dragged on for close to two decades serving certain enterprise and HPC customers.



View at TechPowerUp Main Site

 

[londiste]

Last generation was Itanium 9700 "released" in 2017. Same old 8 cores at 32nm. Previous generation was Itanium 9500 released in 2012. Itanium in itself was not bad and at one time it was reasonably well performing solution for the niche it was deployed in. Wider adoption was not expected, at least not at first. Itanium, after all, was a brand new architecture along with its own instruction set - IA-64.

This is something to ponder on - Intel with all its inertia and 800-pound gorilla might was not able to successfully push out a new instruction set to enterprise.
There were HP Itanium servers, there was software - obviously Linux support but also Windows Server and Microsoft even created things like Visual Studio and SQL Server for it.
Itanium's main downfall was x86-64...

 

[Vya Domus]

Itanium was on life support for ages. Just like it was the case with Xeon Phi, they failed to grasp what the industry needed.

 

[HTC]

Am i the only one that read that title as "Intel Officially sinks the Titanic"?

On a more serious note, i didn't even know Itanium was still being made: i thought it had been "retired" long ago.

 

[eidairaman1]

I guess future iterations will be Xeons

 

[bajs11]

think the first time I heard of itanium was some 15yrs ago then not much at all
and now all of the sudden its declared dead...
its like an elusive rare animal that people knew about but then assumed extinct and then many years later officially declared extinct

 

[R0H1T]

(T)itanic - there she goes!

Am i the only one that read that title as "Intel Officially sinks the Titanic"?

Nope

 

[bajs11]

(T)itanic - there she goes!
Nope

they must have managed to make a 1picometer cpu and predicted the future of an alternate universe in which intel went out of business because a certain individual aboard the Titanic didn't go down with the ship and later on became the ceo of Cyrix that created much better cpus and put intel out of business

this should be the plot for Titanic 2

 

[ShurikN]

Wanted to say "So long, you shall be missed", but it never reached general public so there wouldn't be anything to miss

 

[trparky]

Itanium's main downfall was x86-64...

Pretty much this. There's far too much legacy software in the industry that can't just be recompiled to work on Itanium. Sure, x86 can be emulated on Itanium but performance sucked.

 

[efikkan]

Itanium's main downfall was x86-64...

Pretty much this. There's far too much legacy software in the industry that can't just be recompiled to work on Itanium.

Itanium's downfall was the fact that it didn't scale well compared to Athlon64 among others. It needed a tremendous amount of resources to keep up with more efficient superscalar x86 implementations.

Itanium remains a good example of overengineering; a product which may sound good on paper, but performs poorly in reality.

 

[trparky]

Itanium's downfall was the fact that it didn't scale well compared to Athlon64 among others.

A lot of that came from the fact that the compilers had to do all the hard work. Little do people know but in your common x86-64 chip there's a lot of optimization of the CPU instructions going on behind the scenes at the silicon level before even one instruction is executed. There was none of that happening with Itanium, all of that had to be done at the compiler level which they generally weren't able to do.

 

[efikkan]

A lot of that came from the fact that the compilers had to do all the hard work. Little do people know but in your common x86-64 chip there's a lot of optimization of the CPU instructions going on behind the scenes at the silicon level before even one instruction is executed. There was none of that happening with Itanium, all of that had to be done at the compiler level which they generally weren't able to do.

Not really. First of all, Itanium is a very different ISA, and requires completely rewritten code. All the runtime optimizations and compiler optimizations in the world could never help that. But despite this problem, even the software written for Itanium didn't scale well, so it wouldn't have mattered even if they had more adpotion.

 

[trparky]

requires completely rewritten code.

I wouldn't go so far as to say "completely rewritten code" because in the end C(++) is C(++), it doesn't matter. It's the compilers that make the magic really happen, it's what transforms the human readable code into machine code. Without the compilers and the optimizations that those compilers do it doesn't matter what ISA you have, your code will run like shit. x86-64 has the benefit that pretty much everyone knows how to optimize for it, not only that but there's a lot of architecture optimizations that are being done at the silicon level.

 

[efikkan]

I wouldn't go so far as to say "completely rewritten code" because in the end C(++) is C(++), it doesn't matter. It's the compilers that make the magic really happen, it's what transforms the human readable code into machine code. Without the compilers and the optimizations that those compilers do it doesn't matter what ISA you have, your code will run like shit. x86-64 has the benefit that pretty much everyone knows how to optimize for it, not only that but there's a lot of architecture optimizations that are being done at the silicon level.

Actually not. Itanium features explicit parallel instructions, and compilers are limited to working with just a few instructions within a scope, there is no way a compiler could be able to properly structure the code and memory to leverage this. It's kind of similar to SIMD(like AVX), the compiler can vectorize small patterns of a few instructions, but can never restructure larger code, so if you want proper utilization of SIMD you need to use intrinsics which are basically mapped directly to assembly. No compiler will ever be able to do this automatically.

 

[jabbadap]

Hrh. So at last HP gave up with that, good riddance I guess.

 

[BadFrog]

I heard a few years ago that Intel wanted to EOL it for years but couldn’t due to contracts with giants like HP who had their databases running these cpus and would cost them billions to switch over. I guess the contract is finally up? ¯\_(ツ)_/¯

 

[moproblems99]

in the end C(++) is C(++), it doesn't matter

Not really. Although I am little rusty on my C, I do not believe system calls are abstracted in libc or the STL in C++. Sytem calls very much depend on the system.

Edit: Disregard, I misread your statement.

 

[jabbadap]

I heard a few years ago that Intel wanted to EOL it for years but couldn’t due to contracts with giants like HP who had their databases running these cpus and would cost them billions to switch over. I guess the contract is finally up? ¯\_(ツ)_/¯

It's ended in year 2017 if I remember correctly. But that's not all of it, HP actualy paid intel to continue manufacturing them, intel could have declined.

 

[FordGT90Concept]

Sad. Intel should have tried to sell Apple Itaniums for Mac instead of Core and Xeon. There was no backwards compatibility argument there and Intel could have evolved the product to be more attractive to consumers with Apple's revenue. I don't like that it is being axed because that's another competitor gone but I understand why they are.

 

[TheinsanegamerN]

Sad. Intel should have tried to sell Apple Itaniums for Mac instead of Core and Xeon. There was no backwards compatibility argument there and Intel could have evolved the product to be more attractive to consumers with Apple's revenue. I don't like that's being axed because that's another competitor gone but I understand why they are.

Itanium is just as much a "competitor" for CPUs as VIA is right now, AKA completely worthless.

 

[moproblems99]

because that's another competitor gone

Who are they a competitor to?

 

[Deleted member 158293]

I guess their last support contract ended.

 

[efikkan]

Sad. Intel should have tried to sell Apple Itaniums for Mac instead of Core and Xeon. There was no backwards compatibility argument there and Intel could have evolved the product to be more attractive to consumers with Apple's revenue. I don't like that's being axed because that's another competitor gone but I understand why they are.

I don't recall them having any Itanium version even close to efficient enough to consider that.

Who are they a competitor to?

x86, Intel and AMD
Power, IBM (alive, but barely)
Sparc, Sun (dead)
Alpha, DEC (dead)
PA-RISC, HP (dead)
MIPS (really dead, no market share, but other companies keep MIPS variants still in "active" development)

 

[moproblems99]

ah, so competitor in name only, not really a competitor