Alfonso Maruccia

The big picture: The x86-64 instruction set was initially announced by AMD in 1999, providing a significant upgrade to the leading PC computing architecture. The technology proved highly successful, and Intel had to chase after a competitor for the first time in x86 history. Things could have been much different, though.

Intel had a solution ready to add 64-bit features to the "classic" 32-bit x86 ISA, but the company chose to push forward with the Itanium operation instead. A new snippet of technology history has recently emerged from a year-old Quora discussion. Intel's former "chief x86 architect," Bob Colwell, provides a fascinating tidbit of previously unknown information.

AMD engineer Phil Park was researching the history behind the x86-64 transition, when he discovered the conversation. Colwell revealed that Intel had an inactive internal version of the x86-64 ISA embedded in Pentium 4 chips. The company's management forced the engineering team to "fuse off" the features.

The functionality was there, but users could not access it. Intel decided to focus on the 64-bit native architecture developed for Itanium instead of x86-64. The company felt that a 64-bit Pentium 4 would have damaged Itanium's chances to win the PC market. Management allegedly told Colwell "not once, but twice" to stop going on about 64-bits on x86 if he wanted to keep his job.

The engineer decided to compromise, leaving the logic gates related to x86-64 features "hidden" in the hardware design. Colwell bet that Intel would need to chase after AMD and quickly implement its version of the x86-64 ISA, and he was right. Itanium CPUs had no native backward compatibility with 16-bit and 32-bit x86 software, so the architecture was one of the worst commercial (and technology) failures in Intel's history.

Also see: The Rise, Fall and Renaissance of AMD

The x86-64 ISA was fully compatible with "legacy" x86 code while introducing a new 64-bit mode with more powerful instructions, larger vector registers, and more. It also provided the ability to use vastly larger virtual and physical memory pools compared to 32-bit CPUs. AMD first implemented the x86-64 instruction set with K8-based Athlon 64 and Opteron chips, which ultimately forced Intel to "go 64-bit" with a modified version of the NetBurst architecture (Pentium 4).

Bob Colwell made significant contributions to Intel's history, managing the development of popular PC CPUs such as Pentium Pro, Pentium II, Pentium III, and Pentium 4 before retiring in 2000. Meanwhile, today's x86 chips marketed by Intel and AMD still retain full backward hardware compatibility with nearly every program developed for the x86 architecture.

Permalink to story:

Intel could have beaten AMD to 64-bit transition but wrongly chose not to, engineers reveal

yRaz

Itanium was an interesting platform but looking back, the writing was on the wall for 64bit and I feel it should have come about 2 years earlier. Intel missing the boat was no only a missed opportunity for them, but it slowed the adoption of 64bit processing industry wide. I remember in school, we went with the powerPC G5s for their 64 support.

AIC1Drew

Had just built a PC in late 2005 with an AMD Athlon 64 X2 4800+.

Axeia

Man that was one costly mistake they could have prevented. I remember news from back then saying Microsoft told intel to do it AMDs way forcing Intel into having to license x86_64 from AMD.

If AMD wasn't first to market then it would have been the other way around.

Itanium never moved out of the server market and even there only lived a relatively short live. Opteron was much cheaper and performant enough for most use cases.

antiproduct

I remember back in the day when we called it the "Itanic" processor.

bviktor

yRaz said:
Itanium was an interesting platform but looking back, the writing was on the wall for 64bit and I feel it should have come about 2 years earlier. Intel missing the boat was no only a missed opportunity for them, but it slowed the adoption of 64bit processing industry wide. I remember in school, we went with the powerPC G5s for their 64 support.

Not a very good use of terminology. Itanium was also 64-bit. You meant x86-64.

x86 isn't very good, that's why Intel tried to dump it. But they failed, because it's pretty darn hard to get rid of decades of legacy.

Arm is doing that, very slowly... it's been decades as well, but still only maybe halfway there to get rid of x86 completely.

In any case, we all won with AMD beating Intel to x86-64, because this way they ended up cross-licensing, I.e. AMD licensed x86 from Intel, and Intel ended up licensing x86-64 from AMD. Which ensured that we, the end users, will keep seeing at least some competition in this space.

yRaz

bviktor said:
Not a very good use of terminology. Itanium was also 64-bit. You meant x86-64.

x86 isn't very good, that's why Intel tried to dump it. But they failed, because it's pretty darn hard to get rid of decades of legacy.

Arm is doing that, very slowly... it's been decades as well, but still only maybe halfway there to get rid of x86 completely.

In any case, we all won with AMD beating Intel to x86-64, because this way they ended up cross-licensing, I.e. AMD licensed x86 from Intel, and Intel ended up licensing x86-64 from AMD. Which ensured that we, the end users, will keep seeing at least some competition in this space.

ARM will never replace X86 for the same reason IPv6 will never fully replace IPv4. Too much legacy hardware and software uses it.

And what I meant with Itanium, yes I used the term poorly, is that while it was cool we all lost because they should have tried to bring 64bit support to consumers more quickly. I still remember the "4GB" of ram hack people were using to skimp around the 3.4gig limit and lots of games ended up being 32bit instead of 64 because of it. There are legacy games where mods are limited by that 3.4GB limit at the source code level. Oblivion is a big one.

maxxcool7421

Well they sunk so much money into Itanium they had zero choice

Guillermo Belli

I remember the first Pentium 4 CPUs with 64 bit support were slower when operating in 64bit mode than in 32bit mode.

Theinsanegamer

bviktor said:
Not a very good use of terminology. Itanium was also 64-bit. You meant x86-64.

x86 isn't very good, that's why Intel tried to dump it. But they failed, because it's pretty darn hard to get rid of decades of legacy.

Arm is doing that, very slowly... it's been decades as well, but still only maybe halfway there to get rid of x86 completely.

In any case, we all won with AMD beating Intel to x86-64, because this way they ended up cross-licensing, I.e. AMD licensed x86 from Intel, and Intel ended up licensing x86-64 from AMD. Which ensured that we, the end users, will keep seeing at least some competition in this space.

X86 is good enough it has dominated every market short of ultra mobile.

It's very flexible in it's design, so you can have small efficient cores or scale up to things like zen 5.

Apple made high performance ARM, using a metric shitload of transistors (more then twice as many as a 7950x) and requires optimization to run effectively.

Guillermo Belli said:
I remember the first Pentium 4 CPUs with 64 bit support were slower when operating in 64bit mode than in 32bit mode.

That's because netburst was never designed to be 64 bit in the first place.

Posts: 1,375 +419

yRaz

Posts: 6,950 +10,778

AIC1Drew

Posts: 82 +77

Axeia

Posts: 211 +252

antiproduct

Posts: 347 +519

bviktor

Posts: 1,688 +2,422

yRaz

Posts: 6,950 +10,778

maxxcool7421

Posts: 483 +719

Guillermo Belli

Posts: 89 +115

Theinsanegamer

Posts: 5,888 +11,211

Similar threads