When Intel struggled with its 10nm process technology a few years ago, some investors suggested that the company would be better-off spinning its chip production into an independent foundry, leaving the core of the company to focus on chip design instead. Bucking these calls, however, Intel opted to keep chipmaking in-house, even going as far as to creating Intel Foundry Services to use those facilities to do contract chipmaking for other chip designers.

With the significant capital required to scale up the chip fab side of the business, it's a decision that, even today, Intel executives still get asked about. That was once again the case yesterday, at Intel's investor-focused AI Everywhere event at the Nasdaq MarketSite, where Intel CEO Pat Gelsinger reiterated that the company is not going to spin off their foundries.

"The idea of the internal foundry model, we think, is the right path for us in the current environment," Gelsinger told Reuters.

IFS is currently a distinct manufacturing operations unit within Intel that operates like 'an internal foundry', which the company then 'outsources' production of its processors and other products. Since returning to Intel, Gelsinger has been steadfast about wanting IFS to stay that way, keeping IFS an internal unit rather than to spin it off. It's a decision that's been in notable contrast to some other Intel divisions, such as Mobileye and the Programmable Solutions Group, which have been (or will be) spun off into separate businesses.

With that said, Intel will be bringing more transparency to the financials of its foundry division. Starting from Q2 next year, Intel will report financial results of IFS as if it was a separate business, which will give a clear understanding how much the unit earns and provide a better understanding of how IFS operations stack up against those of TSMC, Samsung Foundry, GlobalFoundries and other top contract chipmakers.

Ultimately, Intel believes that there are clear benefits to operating in a unified manner, especially, as explained by Gelsinger in his interview, that Intel is using the majority of the factory's capacity right now.

Source: Reuters

Comments Locked


View All Comments

  • ballsystemlord - Friday, December 15, 2023 - link

    Now if only we knew what would become of their (arc) dedicated GPU division...
  • tipoo - Tuesday, December 19, 2023 - link

    They also just reiterated Battlemage is coming out and moving higher up
  • ballsystemlord - Wednesday, December 20, 2023 - link

  • evanh - Friday, December 15, 2023 - link

    Wow! Some weed smoking a happening! The fabrication facilities heavily rely on the x86 juggernaut to keep it viable. If Intel was to split itself and then contract out all its x86 business to the lowest bidder then IFS would fall away like GF has.
  • lmcd - Saturday, December 16, 2023 - link

    That is a weird statement. GloFo was 4th to a 14nm-class node, and was on track to be 4th to a 14nm-class node for at least half a decade. Two seconds on Wikipedia shows GloFo "second sourced" their 22nm and 14nm, one of which came from STMicro (and went to Samsung first) and the other of which came from Samsung.

    Intel is either 2nd or 3rd, with a pretty compelling 4nm launch demonstrating comparable density from their "4nm" performance libraries to Samsung's average across all of their 3nm libraries.

    Those aren't remotely the same thing. If Samsung lays another egg, we could legitimately see Nvidia build its consumer product on Intel 1.8nm.
  • zamroni - Tuesday, December 19, 2023 - link

    do you mean fake 1.8nm?
    intel even use tsmc n5 and n6, instead of their fake 4, to produce the gpu, soc and io tiles of meteor lame.
  • lmcd - Friday, December 22, 2023 - link

    Intel's 4 isn't fake, but it is incomplete. We'll see what their 3nm does to finish the node. Intel 7 was a very good DUV node but obviously TSMC's EUV 7nm and refined 6nm outclassed it. Samsung's 8nm was easily as bad though, and while they have less wins total compared to 8nm (in part because TSMC scaled up 5nm quite well), they still have customers.

    Intel's IO tiles and SoC tiles were TSMC 6 because TSMC 6 is strictly superior to Intel 7 in lower-power scenarios. Intel 4 literally doesn't have low power libraries, so that was off the table too.

    Intel's GPU was supposedly originally going to be TSMC 3nm, which Intel fully admits it doesn't have an answer to right now. The tile instead got backported to 5nm (at least that's what I read) and was part of why Meteor Lake slipped. Porting from TSMC to Intel is not trivial at this time, and Intel 4 is low enough volume right now that there's no way that work was getting done.

    All of this to say -- Intel pragmatically went with the market leader for substantial parts of their newest CPU. None of that changes that they're still not in 4th place, and they're arguably tied for 2nd given how poor Samsung's recent track record has been. Samsung's early GAA attempts on 3nm might end up being their Intel 10nm patterning.

    Nothing in Intel 4 is all that ambitious, and the same can mostly be said for Intel 3. Intel 1.8, again we'll see, but "neck and neck for 2nd place" is a reasonable place to be. You can make a lot of money in 2nd place. 4th place starts to get into commodity territory.
  • drwho9437 - Sunday, December 17, 2023 - link

    You are delusional. Intel may well be first to gate all around nodes mass production. TSMC has been in the lead for just a few years compared to the sweep of 50+ years Intel has been around...

    Intel fell behind AMD in the P4 era, then it was better than AMD, then it was worse. I think people who say things like this haven't been around long enough to even remember. I suspect you weren't even alive in the 90s or at most were a small child.
  • evanh - Sunday, December 17, 2023 - link

    The PC made Intel, not the other way round.
  • FunBunny2 - Wednesday, December 20, 2023 - link

    legend has it (I recall reading the quote, but can't find it right now) that the Intel CEO at the time had this to say:
    I'd rather have the 8086 in every Ford than every PC.

    which makes sense, if one believed the ~2,500 sales expectations of IBM (and no clone yet considered).

    and it wasn't the PC that made itself: it was 1-2-3 which was written to MS/DOS assembler alone. Mitch made both Uncle Bill and Mr. Blue rich

Log in

Don't have an account? Sign up now