How TSMC beat Intel to be Apple's main chip foundry

Jump to First Reply
Posted:
in Current Mac Hardware

TSMC and Apple's partnership hasn't always been certain, with a new interview detailing how the company fended off an attempt by Intel to become Apple's chip foundry partner in 2011.

Elderly man with glasses and grey sweater sitting on a sofa, speaking into a microphone, surrounded by framed photos and awards on wooden shelves.
TSMC founder Morris Chang - Image Credit: Acquired/YouTube



TSMC has a long history being Apple's chip producer, with it making everything from iPhone chips to Apple Silicon. However, there was a chance that Intel could've been in that role.

In an interview with YouTube channel Acquired, TSMC founder Morris Chang explained how Apple paused talks with TSMC in February 2011, to hear a proposal from Intel. At the time, Intel wanted to become the firm to construct Apple's chips, in the midst of when Intel chips were used throughout the Mac lineup.

Despite having a history with Intel, it apparently only took two months for Apple CEO Tim Cook to decide to use TSMC instead of Intel.



Chang recounts that a private meeting with Cook in March of that year effectively sealed the deal. He traveled to Apple's headquarters to talk to Cook about the pause in discussions, only to be reassured by him.

Cook reportedly told Chang "Intel just does not know how to be a foundry."

Communication is key



Talking about the relationship with Apple, Chang boasts that he "wasn't too worried" about Intel's offer, due to TSMC's manufacturing capabilities and how it handles its clients.

"I knew a lot of Intel's customers in Taiwan, and none of them liked Intel," he explains. "Intel always acted like they were the only guy for microprocessors."

Chang also brought up that TSMC as a foundry business "does not compete with customers," meaning that even if Intel worked in good faith, "they do have a the conflict of interests."

TSMC's approach was also one where it was responsive to the whims of the clients. "When the customer asks a lot of things, we have learned to respond to every request," Chang states.

"Some of them were crazy, some of them were irrational, but we respond to each request courteously," he continued. Intel has never done that."

An example of this was early on in the relationship, with Apple causing major changes to the TSMC chip roadmap in 2014. At a time when TSMC wanted to shift from a 28-nanometer planar process to a 16-nanometer FinFET version, Apple decided it wanted to use a custom 20-nanometer planar node instead.

With a lack of research and development teams at the time, TSMC couldn't create two different processing technologies at the same time. In the end, TSMC decided to keep Apple happy by shifting all of its development to what the iPhone maker wanted.

The approach worked, as Apple went from a dual-sourced approach with TSMC and Samsung for the A8 and A9 chips, in favor of a TSMC-only chip supply chain.

Over the years, Intel has been keen to win back Apple as a client. So far, Apple has not entertained the prospect at all.



Read on AppleInsider

sconosciuto

Comments

  • Reply 1 of 9
    blastdoorblastdoor Posts: 3,669member
    I wonder if apple ever considered Global Foundries. Back then, Global Foundries was closer to the leading edge than they are today. 
    watto_cobra
     1Like 0Dislikes 0Informatives
  • Reply 2 of 9
    Enjoyed the backstory but really, I doubt Intel was ever seriously in consideration. Their arrogance with respect to customer requests is completely unsurprising.
    watto_cobradanox
     2Likes 0Dislikes 0Informatives
  • Reply 3 of 9
    blastdoorblastdoor Posts: 3,669member
    Enjoyed the backstory but really, I doubt Intel was ever seriously in consideration. Their arrogance with respect to customer requests is completely unsurprising.
    We know that they were in serious contention back in 2007. Back then intel was the undisputed leader in manufacturing process and it was inconceivable at the time that anyone would ever surpass them. 

    But yeah — they were arrogant and that arrogance has destroyed them (along with the shortsighted greed of executives and the board). 
    netroxwatto_cobra
     2Likes 0Dislikes 0Informatives
  • Reply 4 of 9
    ajmasajmas Posts: 602member
    Sounds like the only way for Intel to even to try to win Apple, or any other boundary would be to split. This would result in Intel becoming a chip design business only, while spinning off its foundry to a separate ISMC (Intel Semi-Conductor Manufacturing) or USSMC (US Semi-Conductor Manufacturing) business?

    Ironically, the business attitude at Intel sounds very similar to that of other US businesses, such as Boeing and John Deer, which is hurting them.
    netroxwatto_cobradewme
     3Likes 0Dislikes 0Informatives
  • Reply 5 of 9
    And Chang has said, TSMC has a “Night Hawk Team” to compete with Samsung on 16nm order from Apple, in those days, TSMC’s engineers work in 3 Shifts for 24Hrs to catch up and win Samsung eventually.
    watto_cobradewme
     1Like 0Dislikes 1Informative
  • Reply 6 of 9
    blastdoorblastdoor Posts: 3,669member
    ajmas said:
    Sounds like the only way for Intel to even to try to win Apple, or any other boundary would be to split. This would result in Intel becoming a chip design business only, while spinning off its foundry to a separate ISMC (Intel Semi-Conductor Manufacturing) or USSMC (US Semi-Conductor Manufacturing) business?

    Ironically, the business attitude at Intel sounds very similar to that of other US businesses, such as Boeing and John Deer, which is hurting them.
    I’d frame it as Intel becoming a foundry only and spinning off the chip design business. I think the design business is going to be less profitable long term than manufacturing. I know that’s contrary to conventional wisdom, but I think times are changing. 
    watto_cobraajmas
     2Likes 0Dislikes 0Informatives
  • Reply 7 of 9
    Let’s face it.
    No technology business model is capable to handle design/sell/use her own chips, and run a fab to produce it at the same time, no, you just can’t have both.
    These 2 business need tremendous capital, talented engineers, successful products market to support, to keep it run, non-stop for decade.
    A flopped product will choke the fab, but a fab has to run in economic scale, keep investing it to play catch up, 28nm, 20nm, 16nm..7nm…5nm..2nm.
    Intel has had all the resource to run its own fab, just unable to stay float. No one steal anything from anyone, just winner and loser in a competition, it is a world cup, even Huawei has mimic 7nm production to survive, but Intel can buy all equipments from AMSL and hire people around the world, and consider about its market share, how come will Intel lose the game?
    Thinking of you are a automobile manufacturer, you can design your own new model, a new engine, the interior, select your material combination, but you can’t run your own steel plant, tire plant, plastic parts plants, etc.
    How many car your sell each year to support all those business? And now we are talking about one of the most advance and sophisticated technology product production in the world, USD$30B for one fab, it is 2nm, hundreds billion transistors in a cookie size chip.

     0Likes 0Dislikes 0Informatives
  • Reply 8 of 9
    danoxdanox Posts: 3,510member
    raymondai said:
    Let’s face it.
    No technology business model is capable to handle design/sell/use her own chips, and run a fab to produce it at the same time, no, you just can’t have both.
    These 2 business need tremendous capital, talented engineers, successful products market to support, to keep it run, non-stop for decade.
    A flopped product will choke the fab, but a fab has to run in economic scale, keep investing it to play catch up, 28nm, 20nm, 16nm..7nm…5nm..2nm.
    Intel has had all the resource to run its own fab, just unable to stay float. No one steal anything from anyone, just winner and loser in a competition, it is a world cup, even Huawei has mimic 7nm production to survive, but Intel can buy all equipments from AMSL and hire people around the world, and consider about its market share, how come will Intel lose the game?
    Thinking of you are a automobile manufacturer, you can design your own new model, a new engine, the interior, select your material combination, but you can’t run your own steel plant, tire plant, plastic parts plants, etc.
    How many car your sell each year to support all those business? And now we are talking about one of the most advance and sophisticated technology product production in the world, USD$30B for one fab, it is 2nm, hundreds billion transistors in a cookie size chip.


    No technology business model today can take on everything however, as technology advances, you may need to reconsider (DeepSeek in the news currently is causing that now). Apple was forced to reconsider what they could do because Motorola, IBM and Intel said no, for Steve Jobs and Apple, the third time led Apple to do something they didn’t want to do it, but they had no choice but to do it, and now we have Apple Silicon. 

    In addition, Apple servers, AI models and in house Cuda toolkit like software are now on the radar for Apple not because they want to but because if Apple wants to advance forward they have no choice. After all you don’t want to be Xerox, Kodak or Intel.
    edited February 1
    neoncat
     0Likes 1Dislike 0Informatives
  • Reply 9 of 9
    davendaven Posts: 739member
    An amazing part is that American CEOs make a ton of money because they claim to be key to a company’s success yet we can see that isn’t always true.
     0Likes 0Dislikes 0Informatives
Sign In or Register to comment.