Monday, November 16, 2020

Intel’s Disruption Is Now Complete

James Allworth (via Marcelo P. Lima, Hacker News):

Indeed, that deal between Apple and Intel was more important for Intel than it could have ever possibly realized. But it wasn’t because Intel had sewn up the last of the desktop computer processor market. Instead, it was because Intel had just developed a relationship with a company that was thinking about what was coming next. And when Apple were figuring out how to power it — and by it, I’m talking about the iPhone — they came to their new partner, Intel, for first right of refusal to design the chips to do.


Here’s what Otellini decided to do, when presented with the option to power the iPhone:

We ended up not winning it or passing on it, depending on how you want to view it. And the world would have been a lot different if we’d done it,” Otellini told me in a two-hour conversation during his last month at Intel. “The thing you have to remember is that this was before the iPhone was introduced and no one knew what the iPhone would do… At the end of the day, there was a chip that they were interested in that they wanted to pay a certain price for and not a nickel more and that price was below our forecasted cost. I couldn’t see it. It wasn’t one of these things you can make up on volume. And in hindsight, the forecasted cost was wrong and the volume was 100x what anyone thought.


What about this chart is interesting? Well, it turns out, it bears a striking resemblance to one drawn before — actually, 25 years ago. Take a look at this chart drawn by Clayton Christensen, back in 1995 — in his very first article on disruptive innovation.


As someone who worked on Intel’s phone chip: we definitely didn’t win it. We fucked it up twelve ways to Sunday. Why: giant egos. There were turf wars between Austin, Santa Clara and Israel over who would design it, and the team that won out had long since lost its best principle engineers and had no clue how to spin the architecture to meet the design win. Otellini’s hindsight hedge is pure spin: we knew the landing zone, we just didn’t know how to get there. And the aforementioned turf war guaranteed we didn’t get access to other teams’ talent. I’m bitter because it was a really fun team when I moved from Motorola to Intel Austin, and then it just corroded over political battles.

John Gruber:

It remains to be seen if other ARM chip vendors will surpass the x86 platform in performance and efficiency. But it’s starting to look like that’s inevitable — Apple is just far ahead of the pack.

Tony Fadell:

  • ’92 - Started working w/ ARM at General Magic
  • ’01 - Bought ARM back to Apple by choosing SoC w/ Dual ARMs for the iPod
  • ’08 - Solidified ARM as the future of the iPhone & iPad w/ a showdown vs. Intel “Intel is what Steve wants!” was the refrain by my peers then

Mike Dauber:

Bob Mansfield, Jeff Dauber, and Lynn Young were the ASIC leadership team that came over from Raycer Graphics in ’99. Later augmented by PA Semi. I believe Bob convinced Jobs that Apple needed their own ASIC team. He was right.


Update (2020-11-27): Stephen Shankland (via Hacker News):

Losing Apple’s business will sting. New Street Research analyst Pierre Ferragu estimated in a Wednesday report that 4% to 5% of Intel’s revenue comes from Apple. But it’s just one of the concerns Intel will need to address.

Intel said it’s “relentlessly” focused on building leading chips. “We welcome competition because it makes us better,” Intel said in a statement. “We believe that there is a lot of innovation that only Intel can do,” including supplying chips that span the full price range of PCs and that can run older software still common in businesses.

4 Comments RSS · Twitter

The story with about Otellini is a little confusing. When I first read the story years ago, I remember this specific sentence,

>At the end of the day, there was a chip that they were interested in that they wanted to pay a certain price for and not a nickel more and that price was below our forecasted cost.

I didn't read that as the CPU, I read that as a support chip. Do we know if it was the cpu or a support chip?

It's easy to forget how bold a move it was for Apple to invest heavily in CPU/SoC design at that point in time. A lot of people were deeply skeptical about Apple's purchase of Palo Alto Semiconductor, for example. The conventional wisdom was very much along the lines that only Intel and one or two other players had the knowledge and capacity to be competitive. The "smart people" were convinced that systems manufacturers were playing a fool's game to be involved in chip design.

HP was abandoning PA-RISC. Alpha was dead and buried. SGI had disbanded its high performance MIPS team and spun off the remainder of MIPS as a separate company targeting embedded applications. Even Intel's Itanium was sputtering. SPARC would survive (and systems are still sold today) as long as Oracle could continue extracting outrageous support contracts from large enterprises, but no one seriously viewed SPARC as a growth market. Aside from SUN/Oracle's rear guard action and IBM's prestige investment in the POWER architecture, nearly everyone else gave up, jumped on the Intel train, and they were thought wise to do so. Chips for phones were thought to be niche markets -- Intel even made some successful ARM-based chips for phones like the late Palm Treos, but they sold that business off to Marvell.

Kudos to Jobs and others at Apple for their vision. It was not at all the obvious thing to do back then.

@Ryan Collins

The chip in reference is the actual SoC. Remember at the time, SoC wasn't a well known terms. And most people would refer to chip as simply either CPU or GPU. So Otellini's "chip" was in reference to Apple's SoC. With the same GPU tech from IMG which Intel also licensed. ( Along with many other things )

@Josehill : you should read up the story of the 1st ARM chip ever. The people involved thought there must be some good reason so few people made CPUs, and that they were crazy to try. They were sure they would find an insurmountable barrier at each step along the way... until they received first silicon and it just worked.

Leave a Comment