Apple's operating chief Jeff Williams was in Taiwan today to attend the 30th anniversary ceremony of Taiwanese Semiconductor Manufacturing Corporation, more commonly abbreviated as TSMC. There, he spoke about artificial intelligence, the future of the semiconductor industry, and more.
Williams first reflected on how Apple and TSMC began working together seven years ago. Today, TSMC is the exclusive supplier of the A10X Fusion and A11 Bionic processors in the latest iPhones and iPads.
First off, thank you. It's a real honor to be here with this distinguished group, and we're here of course to celebrate TSMC's 30 years. And it's amazing as you've seen in the slides, how far technology has been driven over that time. TSMC got its start shortly after the introduction of the legendary Cray II supercomputer, and 25 years later we put this same processing power in people's pockets with an iPhone 4 in 2010. It really is remarkable, and it was actually 2010 that the first seeds of our partnership between Apple and TSMC were planted.
I had flown to Taiwan and had dinner with Dr. Chang and Sophie at their house. It was a wonderful dinner. We were not doing business with TSMC at the time, but we had a great conversation. We talked about the possibilities of doing stuff together, and we knew the possibilities would be great if we could take leading edge technology and marry it with our ambitions. And what seems obvious now, wasn't then, because the risk was very substantial.
The nature of the way Apple does business is we put all of our energy into our new products, and we launch them, and if we were to bet heavily on TSMC, there would be no backup plan. You can not double plan the kind of volumes that we do. We want leading edge technology, but we want it at established technology kind of volumes, and so that may be want Dr. Chang is referring to when he says "intense."
Williams said Apple and TSMC have gone on to ship over half a billion chips as part of a "wonderful partnership" between the two companies, in which billions of dollars has been invested to fulfill Apple's significant production demands.
…Together we decided to take the bet, take the leap, and Apple decided to have 100 percent of our new iPhone and new iPad chips, application processors, sourced at TSMC. And TSMC invested $9 billion and had 6,000 people working 'round the clock to bring up a Tainan fab in a record 11 months. And, in the end, the execution was flawless. And we've gone on together to ship over half a billion chips together in that short window. And I think TSMC has invested $25 billion. $9 billion on that first venture — there are very few companies in the world that $9 billion in capital across everything, not a single bet. So for that, we thank you, Dr. Chang, and everybody at TSMC. It's been a wonderful partnership.
Williams was asked to describe his vision of the next ten years in silicon, but he chose to reframe the question as "do we have enough processing power in our silicon to match our ambitions?" and answered accordingly.
It's interesting, when we look back a decade ago, the question we had was "do we have enough processing power in our silicon to match our ambitions?" The big challenge we had as we moved into the mobile revolution was this tradeoff between performance and power, and the view at the time is you had to choose — you've got one or the other.
Largely as a result of what the fabless model has done, what TSMC has done, what many people in this room have done, Simon and his organization from ARM — we have reached a point where those tradeoffs are not necessary. We have performance in thermally constrained environments. And so this opens up for the next decade a whole new world. So for the next decade, the question is not so much "Do we have enough processing power to meet our ambitions?" Though we need to keep working, of course we need to drive better lithography — don't slow down! — but I think the question for us is "Do we have the right ambitions to go utilize this technology in front of us?"
Williams went on to say that Apple thinks artificial intelligence and on-device processing will be key to the future of the semiconductor industry. He believes these advances will help to "revolutionize healthcare" for one example.
We at Apple are not concerned about the talk of a slowing semiconductor industry. Not the case at all. We think the potential is huge. We believe strongly in both the cloud side, but the future will be a lot of on-device processing. We believe this is the best way to deliver great features without sacrificing the responsiveness and the privacy and the security. We see in our brand-new A11 Bionic chip, which is made right here at TSMC, every time somebody takes a photo, there's over 100 billion operations. That's just mind-boggling. In a single photo, over 100 billion operations. The potential is limitless.
We put a neural engine on the chip, and I won't repeat some of the things that Jensen shared, but we have the same view and vision of the potential of AI to deliver a much safer and efficient autonomous system. The neural engine on our chip has already enabled Face ID, processed locally. And so we view that the next ten years is about the ambition to do what Simon's daughter is asking for, to make life better. And probably one of the most significant examples of this is our opportunity to use transistor technology advances and power scaling to revolutionize healthcare. We think the industry is ripe for change. We think there is tremendous potential to do on-device computing, to do cloud computing as well, and to take that learning, and through machine learning, deep learning, and ultimately artificial intelligence, change the way healthcare is delivered. And we can't think of anything more significant than this.
So I think the question in front of us is "Do we have the right ambitions, and can we go do this?" And there is no such thing as autonomous innovation. Human beings dream it. Human beings drive it. And sure, we'll have deep learning, but there's not autonomous innovation, so it's up to us, this generation over the next ten years, to take advantage of what is in front of us in the silicon world. We at Apple are really inspired, for those of us who started many years ago on a green monochrome computer screen, we're super inspired with the state we're in, and I'll just say this: If in the next ten years, from a society standpoint, we just do a few "gee-whiz" things like flying car kind of dreams, and then the rest of the time we're using the faster chips to do the same things we're doing faster, we will have squandered one of the biggest opportunities in front of us. I think we're at an inflection point, much like my colleagues, with on-device computing, coupled with the potential of AI, to really, really change the world. And we couldn't be more excited about it at Apple, and thank you for your time.
China's Economic Daily News reported that Williams also met with iPhone assembler Foxconn's chairman Terry Gou, but no further details were shared. An earlier report speculated that iPhone X production issues could be one focus.
TSMC provided a live stream of the event that remains available to watch for those who prefer to listen to Williams' comments.
