SAN JOSE, Calif. - Is Apple Inc. switching foundry partners?
Apple is mulling a plan to outsource the production of its A4 processor and other ARM-based chips to Taiwan Semiconductor Manufacturing Co. Ltd. (TSMC), according to Digitimes. That article can beread here.
At present, Samsung Electronics Co. Ltd. is Apple's foundry partner. There are also reportsthat Apple is sourcing $7.8 billion in parts from Samsung.
But on the other hand, Apple and Samsung are rivals in handsets, tablet PCs and computers. Some think Apple should build its own fab.
Agreed.... this is a love hate relationship. One cannot make long term plans without a second source. This is especially true for something as IP, knowhow and capital intensive as a state of the art semiconductor logic manufacturing facility with 80,000 WPM capacity.
To those that feel Apple needs its own fab I point to the history of Commodore, Digital Equipment, Data General, Sperry Univac etc. and let history speak for itself
Samsung is a friend and foe to Apple. Thus Apple has to win consumers over innovations not semiconductor components. Japanese were good at components but fail to promote beyond that. TSMC is a natural second source to Samsung on A4/5. Sadly there is no one else out there for Apple except Samsung and TSMC. Apple knows about this.
Perhaps Apple is different. They already have substantial vertical integration, from architecture, apps to chip design. Probably fab is next in line. Perhaps Steve Jobs sees threat emanating from Samsung which draws its strength from vertical integration.
I’m having difficulty with this entire thread. The sky is not falling and “Apple jumping ship and moving to TSMC,” is insane.
When the volume of business is in the billions of dollars and the capital needed to supply the parts is in the billions of dollars, nothing moves without protection of IP and investment needed to deliver. It’s just that simple. At the same time we do live in a competitive world and single sourcing is extremely dangerous. To that end I say Apple must explore alternatives to Samsung. Samsung was has been and will continue to be in the phone and computer business for over a decade. Apple knew this going in as did Samsung. Hurray for TSMC, for making the pitch and generating press that it is expanding and hinting it’s on the hunt for Apple business.
Competition is what this business is all about. Anyone that thinks Apple can leave Samsung high and dry also believes in the tooth fairy. At the same time a dual supplier strategy is a must for everyone.
As for Apple buying and operating a fab, that’s outrageous. One would only have to look at what they have done with most of the other manufacturing, it’s outsourced. Sure they do invest in some of their suppliers. Its common practice when you want it special and guaranteed access to capacity.
“Jump Ship,” now way…..
I will agree Apple should focus on its innovation, not those commodity SoC or flash. If they could come up unique requirements, I believe every major SoC vendors will do everything for them to gain their business. And Apple could cherry pick the best. Why bothering even to design its own SoC? I think it is silly decision for Apple to start with.
I concur with the above comments. Apple need to concentrate on their own competitive advantage (design) and leave the fabrication to others. Parting ways with Samsung is logical as the two are increasingly competing in the mobile arena.
Since Samsung is competing with Apple with their own Hummingbird/Orion/Exynos CPU, this is a logical decision. Apple wouldnt want samsung to know what is its next step. BTW samsung hummingbird was originally based on Intrinsity ARM core, which Apple acquired. Since then Apple/samsung would have diverged their paths in CPU development.
I don't agree Apple to have its own fab. Apple only need to focus on product design and probably some chip designs if they are valued to the company. There are plenty of large fab for Apple to choose from so I don't see any good reason for Apple to build its own fab.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.