New Intel CEO Wants to Beat Apple

Posted on January 16, 2021 by Paul Thurrott in Apple, Hardware, Mobile with 33 Comments

Well, Pat Gelsinger was clearly the right person to lead Intel. In his first all-hands meeting with employees, he emphasized the need to beat Apple.

The account of Gelsinger’s first appearance with employees comes via The Oregonian.

“We have to deliver better products to the PC ecosystem than any possible thing that a lifestyle company in Cupertino,” Gelsinger told employees on Thursday. “We have to be that good, in the future.”

To be clear, Gelsinger is referring here to Apple Silicon and the other chipsets advances that the Cupertino consumer electronics giant has made against more established players like AMD, Intel, and Qualcomm. Apple started to transition the Mac off of Intel chips last year and is expected to finalize that transition next year.

Gelsinger has many other related issues to contend with, too, of course. Key among them is a decision that Intel has now delayed regarding whether it should outsource chipset production so that it can move to smaller manufacturing processes. Intel is currently stuck with 10- and 14-nm chip designs while Apple, AMD, and others have moved to smaller and more efficient 5- and 7-nm processes thanks to TSMC and other manufacturers.

“The company has made strong progress on its 7-nm process technology and plans on providing an update when it reports its full fourth-quarter and full-year 2020 results as previously scheduled on January 21, 2021,” an Intel statement notes. It was issued the day before Intel announced Gelsinger as its next CEO.

On that score, Gelsinger hinted to employees that Intel will likely outsource its most advanced chips while it continues to build most of its chips and evolve its own factories to handle smaller manufacturing processes. The faster this happens, the better.

Tagged with

Join the discussion!

BECOME A THURROTT MEMBER:

Don't have a login but want to join the conversation? Become a Thurrott Premium or Basic User to participate

Register
Comments (36)

36 responses to “New Intel CEO Wants to Beat Apple”

  1. SvenJ

    Beat Apple at what? They aren't technically in the same business. Intel sells processors to 3rd parties. Apple does not. It is somewhat embarrassing that Apple seems to be better at Intel's core business than Intel is, but maybe that is from only one viewpoint. Intel needs to produce a product compelling to it's customers, over alternatives it's customers have. They do need to provide a product to their customers that lets them build a product competitive with Apple. The PC industry isn't exactly in danger of Apple taking over.

    • kjb434

      In reply to SvenJ:

      When Apple decides to remove your manufactured component from it's devices, it could spark a trend at more companies deciding to not use the same component.


      AMD has made inroads in the data center business. Add that arm based data centers are on the rise too.

      The market Intel has is being chipped away (pun intended) from several angles. Even Qualcomm if it gets its act together can build chips as good as Apple further sidelining Intel.


      Doesn't mean x86 is going away, but it's role can be much smaller than today.

    • Paul Thurrott

      I assume they meant beat Apple in processors. :) What else would he mean?
    • nine54

      In reply to SvenJ:


      The challenge for Intel seems to be that this horizontal separation of chip design vs. fabrication is leading to greater agility and other competitive advantages for the vendors within that supply chain. Intel needs to disprove this and demonstrate that once it addresses its execution issues, no one is better positioned to deliver market-leading chip solutions.

  2. gadgetenvy

    How does a new CEO help Intel do this technically difficult thing they have been trying to do for years?

    • bkkcanuck

      In reply to gadgetenvy:

      I don't know if it is possible... if they cannot beat TMSC fab process, then I don't think it will be possible to be as efficient (which will translate into performance as well going forward) as an ARM processor. If this 'beating' Apple with x86 is a long term strategy, then Intel is in trouble. If this is merely a short-term strategy while they try to diversify then it could be possible. IMHO, x86 has a built in disadvantage as to stay in the game long ago they effectively implemented the x86 code on a hardware/microcode decoder that converted it into RISCish micro-ops (needed for the purposes of multiple instruction pipelines and parallelism). The question is how long can Intel maintain the cash cow x86 business before they hit a wall in the future.

    • Paul Thurrott

      He's an engineer not a marketer/businessman, for starters. So maybe he starts making good engineering decisions instead of short-term business decisions.
    • Greg Green

      In reply to gadgetenvy:

      As Paul says it’s a matter of emphasis. The new ceo will look for engineering solutions that are profitable, whereas the previous guy looked at financial solutions that were profitable.


      a negative example is Boeing. Once they moved their HQ out of Seattle and put accountants in charge it changed the whole environment. Project managers went from getting visits from engineer/executives who discussed engineering problems to getting calls from accountant/executives who were interested in budgets and deadlines.


      when your boss understands the technical aspects of your job communication and understanding (both ways) are a lot better.

  3. veermaharaj

    I'm reading this whole thing and noticing, Intel's new CEO did not mention the biggest threat to Intel; AMD.

    Apple's sales account what what single digit percent of the PC market including desktops, laptops and servers?


    Their real threat is AMD but they are doing what they always do which is pretend like AMD doesnt exist. That was and still is their playbook for dealing with their only serious x86 competitior. (And don't forget VIA technologies out of China also has an x86 license and has started making CPUs again which are matching 5th gen intel parts from just a little while ago and are picking up steam.)


    AMD is putting more silicon in servers, desktops and laptops than Apple does at all.

    And given the M1 is being pushed exclusively to Apple devices, and it will make it all the more proprietary, there won't be any clear 1:1 benchmark to compare Apple's PCs to anyone else's.


    So why is Intel even trying to do that?


    Misdirection.

    • Greg Green

      In reply to veermaharaj:

      Companies usually don’t mention the other key competitors name, it gives them free advertising.

      • bkkcanuck

        In reply to Greg Green:

        This was not a press event or an event for the public, this was in internal motivational meeting with employees. While it is true that a competitors name would be best not mentioned at public events or press events -- the same is not true of internal employ meetings. I have worked for companies that did not mention their competitors at the public events, but internal employee events meant to motivate us to compete - the same rules do not apply.

    • nine54

      In reply to veermaharaj:


      AMD is successfully competing with Intel in the x86 space, which is Intel's "home turf." Apple represents the broader always-on, always-connected device market in which Intel has struggled to find its footing. Aside from higher growth potential, this market is a threat to x86 as part of a paradigm shift in computing.


      Both markets are important and are not mutually exclusive, and I'd imagine the company intends to compete in both. That said, I agree that from a priority standpoint, it might make sense to shore up your core market first before trying to take space in another market.

  4. Saarek

    Intel is in real trouble, have been for a while now in fact.


    I remember reading an in-depth head to head test of the 2nd Gen Core i7 vs the 8th Gen Core i7 a couple of years ago. The 2nd Gen held up really well and was only about 40% slower than the 8th generation model, despite there being a 7 year gap between the two. +/- 40% gain in 7 years..........


    If we take a look at the Apple A7 released back in 2013 and compare it against the A14 we can see an average performance gain of about 600% across the board. Just think about that, Intel 7 year gain about 40%, Apple CPU gain about 600%!!!!!!!!


    Sure, as the Apple CPU's have matured the performance gains have been less impressive, but they are still posting 20%+ performance gains across the board year after year with their iPhones. If they can do the same with their Mac chips, and we have no reason at the point to presume that they can't, then where does that leave Intel?


    People will rightly point out that Apple won't licence these chips to other people and the Mac base is relatively small 7-8% market share of the PC Market.


    But that simply means that they have a lot of room to grow and that companies like Qualcomm might take the Arm based PC CPU market seriously and point to Apple as the example.


    Intel could still recover from this, but unless they really get their butts into gear they are dead.

    • illuminated

      In reply to Saarek:

      3rd gen to 4th gen was pretty impressive. I still remember super-hot 3rd gen laptops that could not be used on the lap without coolers. 4th gen were way cooler. After that I did no see much improvement from performance perspective. I was checking every year if there was a good intel CPU that could replace my 4th gen i7 desktop CPU but there was no comparable replacement for 5 years.

  5. 2ilent8cho

    The mistake Intel might be making is believing the M1 is Apple's cutting edge chip and thats what they need to gauge themselves against. We have no idea how fast Apple can actually make their chips yet as we have only seen the entry level slow processors so far, and we have as yet to see how big the year on year gains are going to be, but if the A series iPad and iPhone chips are anything to go by they will be serious gains.

    • Greg Green

      In reply to 2ilent8cho:

      We’ve seen entry level fast processors. The M1 Mac air book and Mac mini, the bottom entries, can now compete well with the top entry MacBook as well as mid level iMacs, IMac Pro’s, and even the trash can macs running i5s and i7s. That’s a $1000 notebook and an $800 MUC going against $2000 and $3000 Mac products.

  6. murray judy

    He's got his work cut out for him. But, I'd give an engineer with good leadership abilities, such as defining the competition in terms that everyone can understand, a much better chance of turning Intel around than a CFO that defines goals in terms of financial performance.


    I wish him well.

  7. RobertJasiek

    One of the things Intel misses is an M1-like CPU for x64. The closest Intel has come is a set of CPUs approaching the goal from different directions, such as 7100U (good enough for fanless barebone but still too hot for mobile devices), Atom cores (fanless but still too slow by far in comparison to M1) and recent notebook CPUs (more or less fast enough but by far too hot for fanless in small mobile devices). Amberlake is the next hope but my guess is that it would have to be TSMC 5nm to possibly compete with the M1 in the fast + fanless segment. Unless Intel gets access to that, it will remain years behind in the race with Apple's mobile chip achievements because Intel does not succeed with its 7nm yet and 5nm of its own is a dream at best.

    Intel CPUs are good enough for Surface Pro, barebones and business PCs but do not compete in the mobile and gaming sectors. (I do not know about the server CPU market, except that Intel is too expensive and maybe inefficient for that.)

    Optimistically, Intel has a chance to learn from TSMC, Apple, AMD and Nvidia. If Intel cannot learn from studying their chips, Intel is doomed in a couple of years or becomes the next IBM.

    • Oreo

      In reply to RobertJasiek:

      According to what Intel has released, Amberlake will still be missing plenty of pieces compared to the M1, including cores for machine learning and dsps. Furthermore, the M1 is the slowest ARM-based chip Apple will release, so we do not even know how far behind Intel really is.

  8. madthinus

    Bob Swan early on discussed the need for Intel to move past the need to be a CPU company for PC and they need to focus on being a chip company for more than just PC. This I felt was smart. There is a need for a lot of chips, not just CPU's for desktops and laptops. Looks like Pat is very much not thinking down that line.


    The thing I am looking for is, will he spun the Fab business out.

    • mattbg

      In reply to madthinus:

      Another possibility is to license the processes of others to produce in their own fabs. This has some appeal as well because one of Intel's key differentiators is their manufacturing capacity and it will take time to reproduce that capacity even if they do decide to outsource to outside fabs.


      I guess the decision will come down to how good they feel about their ability to move past 10nm to 7nm and then quickly onto 5nm. They've been saying they have it in hand now but these decisions will indicate the real state of things. After all, Apple must have seen things they didn't like and you'd presume they have access to inside information as a major buyer thinking of jumping ship.


      Like you said, I think putting Bob Swan in charge was perhaps what they needed to do to shore up their existing business and look for other revenue opportunities to provide cover for their process issues. Now it's time to change focus, and it's good that Intel is making it clear that the focus will shift a bit more in the technology direction because their technological image is under fire at the moment.

  9. gregerst

    It really shows you have far Apple has come up and how low Intel has fallen in a few short years. To imagine that Intel would be chasing Apple in the CPU design and implementation space would have been laughable not long ago. Did Apple steal away the best engineers?

  10. codymesh

    these guys have had six years to move beyond skylake...the next six years will make or break intel

  11. will

    Lets check back in 2 years, then at the 5 year mark...

  12. abdulla77

    Before anything, he needs to first of all change company culture.. The perceptions and mentalities within their leadership needs to change. Then it would take years to shape a new Intel with new blood. Satya when he came in had to rethink and refocus their efforts and that required, to some extent, changing the culture in the company. Not impossible to do.. just needs a lot of persistence and focus.

  13. markbyrn

    Around 12 years ago, Nokia disparaged Apple as that fruit company in Cupertino. If Gelsinger wanted to make a war cry for the troops, he should have said we're up against competitors that include an aggressive anti-competitive trillion dollar technology behemoth in Cupertino. Instead he lamely dismisses them as a lifestyle lightweight. Let's hope Intel does better than Nokia did.

  14. tarnishedtinman

    Apple is the wrong enemy for Intel to concentrate on.


    Instead Intel should be focused on AMD and ARM in general. They are the real threat to Intel. Not some company that only accounts for a small percentage of the PC market and will have no incentive to come back to Intel. Apple isn't going to start selling their custom ARM chips to other companies so that damage is contained.


    Instead focus on offerings that will blunt the advances of AMD into the PC market. spend time looking at why PC companies that have traditionally been Intel are starting to offer ARM based alternatives.


    Going after Apple is only designed to fan Intel's ego. Going after stopping AMD and ARM chips would actually fan profits instead of ego.


    Stock holders won't care if Intel's ego is sated at the expense of profits.



Leave a Reply