Apple silicon: What we know about Ibiza, Lobos and Palma

From Wayne Ma’s “Apple’s Road Map for Mac Chips Shows Likely Advantage Over Intel” ($) posted Friday by the Information:

Apple’s plans for its future Mac processors suggest those new chips are likely to easily outperform Intel’s future processors for consumer PCs, previously unreported details about Apple’s road map show. Apple has already begun working on the next two generations of Mac chips, which are expected to succeed the M1—the first Mac processor Apple designed in-house as it began to move away from Intel, according to three people with direct knowledge of the plans.

Apple’s third generation of Mac processors—which go by the code names Ibiza, Lobos and Palma, according to three people with direct knowledge of the projects—look to be an especially big step up from the processors Intel is expected to begin shipping around that time, analysts told The Information…

If Apple succeeds in blowing past Intel in the performance department, it will be a major vindication of its decision to end the 15-year relationship between the companies and to invest heavily in the expansion of its in-house chip design team. For over a decade, the company’s silicon engineering group has been designing the processors that go into iPhones and iPads, the A-series of chips. Last November it introduced the M1 in some MacBook models, and last month it put improved M1 chips in more of its product line, eliminating Intel processors from all but a couple remaining Macs.

Silicon engineers have a number of technical tricks they can perform to boost the performance of processors, all of which come with trade-offs such as increased power consumption and heat, as well as presenting thorny new design and manufacturing challenges. Designers can increase the number of dies—the hearts of the processor—on the chips, which contain dense thickets of transistors. A single die can contain multiple cores, allowing more tasks to be performed at once.

Silicon engineers can also increase the frequency of chips so they can execute more instructions per second. And they can use more-advanced manufacturing processes to shrink the size of transistors—measured in nanometers—on a die, enabling more-advanced chip designs.

In the case of Apple, the company’s first generation of Mac processors—the M1, M1 Pro and M1 Max—each has one die and is manufactured using a 5-nanometer process. But Apple’s road map calls for steady improvements on those features.

For its second-generation processors, Apple plans to manufacture them using an upgraded version of the 5-nanometer process, two people with knowledge of its plans said. One of the people said the chips will contain two dies. Apple plans to take a much bigger leap with its third-generation processors, some of which will have four dies made using a 3-nanometer process, that person said.

Apple’s fastest processors currently contain 10 compute cores on a single die, which could translate into as many as 40 compute cores for a chip with four die.

My take: They had me right up to 4 dies X 10 cores/die = 40 cores. The math can’t be that simple.

Note: Apple famously code-names its silicon after islands — from humble Staten to colossus Rhodes. Ibiza, Lobos and Palma happen to be Spanish islands — two Canaries and a Mediterranean.

13 Comments

  1. Robert Paul Leitao said:
    Successfully producing chips with sufficient yields while increasing the number of functional cores at these small sizes (3nm and 5nm) are not easy accomplishments. I don’t expect Intel to be competitive with Apple’s silicon on performance anytime soon. Becoming competitive may require Intel to also outsource to an independent foundry. Apple currently has a conspicuous performance advantage over Intel in the desktop PC and laptop markets. This advantage will reshape the PC market over the next few years. The only limit for Apple’s Mac line is the ongoing migration away from traditional PC forms to wearables and other devices. Of course, Apple is already a global leader is wearable devices.

    6
    November 5, 2021
  2. Fred Stein said:
    This is part of big big story. Apple has tossed Samsung, Intel, and soon Qualcomm. Each of them are the classic incumbents being disrupted. NVIDIA’s PC franchise will give way to Macs over a long PC-to-Mac (and iPad/iPhone) conversion.

    It’s much bigger than cost, performance, power consumption, control, and not letting your competitors be your suppliers.

    In wearables, it’s already Apple and the distant pack, which will recede further and further. In the high-end compute workloads, software makers will migrate to Mac. And new software that we can only imagine will land on M-series.

    7
    November 5, 2021
    • Gregg Thurman said:
      In the high-end compute workloads, software makers will migrate to Mac. And new software that we can only imagine will land on M-series.

      Whoop, there it is.

      0
      November 6, 2021
  3. Robert Paul Leitao said:
    For too long Apple has been limited by Intel’s chip designs and delivery timetables. While Apple has taken the strategic and financial risk of designing its own chips, the company is no longer paying the “Intel tax.” Using Apple’s own silicon, the company can bring innovation to market more quickly, reduce the marginal cost of each additional unit manufactured, and increase the integration of hardware, software, services and other solutions across device lines. Apple owns its destiny.

    6
    November 5, 2021
  4. Jerry Doyle said:
    And they say Apple can’t and hasn’t innovated? Innovate my a##! 🙂

    5
    November 6, 2021
  5. Kirk DeBernardi said:
    Stickin’.

    To.

    The.

    Knittin’.

    1
    November 6, 2021
    • Kirk DeBernardi said:
      As I read this, I can’t help but think back to when I first had the notion to invest in AAPL.

      I remember years ago (pre-iPod) an interviewer asked Steve Jobs for a comment about his take on the Mac’s paltry market share versus the Win-Tel gauntlet. Steve replied that he compared it to BMW’s then market share of the automobile industry where, although nowhere near having a strong share of their market, they nonetheless held their ground and created beautifully engineered products that people loved. Then he said that should Apple ever take just a wee bit more market share, it would represent millions (billions?) of dollars for the company.

      I always held the dynamic investment promise of that comment in the back of my mind that should Apple ever come up with something where they could double or triple their market share that it would be a great thing and a great investment.

      Shortly after that comment soon came the iPod > iMac > iPhone gaunlet and eventually Apple’s market valuation topping Microsoft’s.

      I initially bought in small after Steve’s BMW comment as I had little money then, but later bought in heavily at the early ascension of the iPhone.

      Now it’s “déjà vu all over again” and Apple sits poised to disrupt the landscape once more with their hot silicon.

      What a long, strange trip it’s been.

      Never say never and never give up hope.

      4
      November 6, 2021
  6. Gregg Thurman said:
    Apple has already begun working on the next two generations of Mac chips,

    The author lost me here with this great big DUH statement, and again with this one which are expected to succeed the M1. Of course there’s going to be a successor, and Apple has most likely been working on the successor to the M1 for the last 2 to 4 years.

    Nothing new here, just a clever recap of others’ work, repackaged to appear original.

    5
    November 6, 2021
  7. Miguel Ancira said:
    I was not planning on selling until AAPL reached 300. Maybe it is time to revisit that number.

    0
    November 6, 2021
    • Gregg Thurman said:
      I was not planning on selling until AAPL reached 300. Maybe it is time to revisit that number.

      You could sell way out of the money weekly Calls against your shares. The further you go out the lower the probability your shares will get called away. $5.00 out of the money should suffice for the next month or so, but $7.50 out of the money would be a lock. Even then, unexpected good news can drive AAPL up quickly and substantially. To help insure your shares are Called away, try to sell your Calls near what you believe is Monday’s high. I always do my trading of weeklies on Monday for two reasons: weekend news in other parts of the world can be disruptive and, Time premium is highest on Monday for weekly contracts. When selling option time premium is your friend.

      Depending on how many shares you have you can create a very nice supplemental income source without relinquishing your shares.

      1
      November 6, 2021
      • Miguel Ancira said:
        tks, Gregg…I will look into it

        0
        November 6, 2021
      • Lalit Jagtap said:
        Thank you Gregg for sharing your insights about highest time premium on Monday. The Covered Calls are good way to generate additional dividends.

        0
        November 8, 2021
  8. Apple is without a doubt Taiwan Semiconductor’s largest customer. + TSM has far & above the best fabs.+ ASML’s best customer for lithography machines is ‘located in Taiwan.’ Still, as comments above this one note, the entire process is riven with potential pitfalls & millions in early waste. Apple silicon is their most strategic new product since the iPhone, I’ve been told & do believe. Maybe Cupertino can retrofit the old Intel fabs to meet demand for their SoCs.
    Look closely at the latest 10-K, you’ll find fab costs there.

    0
    November 6, 2021

Leave a Reply