Blog Post

Apple’s Billions and Billions

Yesterday, the New York Times published an article examining what it refers to as the upcoming “war” between computer chip manufacturers. It’s an interesting read if you’re desperately into that sort of thing, but what’s most compelling is the assertion that Apple (s aapl) probably invested at least a billion dollars in the iPad’s custom silicon.

As we reported here, Apple bought chip manufacturer P.A. Semi back in April 2008 for a cool $278 million, ostensibly to acquire the company’s engineering talent and manufacturing expertise, and, perhaps, the use of its existing facilities to produce its own custom-designed chips. Perhaps this helped save Apple a little money up-front, if the NYT’s is correct about the development costs of the chips alone;

Even without the direct investment of a factory, it can cost […] about $1 billion to create a smartphone chip from scratch.

Does this mean Apple saved a cool seven hundred million dollars when it bought P.A. Semi? If you’re a company with almost forty billion dollars in the bank, finding the ready cash to develop your own groundbreaking processor doesn’t seem quite such a mammoth undertaking. And I’ll be the first to admit I’m likely oversimplifying the whole thing, but y’know, that Jobs fellow is a wily old fox…

(Chip) War is Hell

From the New York Times;

Now, the chip wars are about to become even more bloody. In this next phase, the manufacturers will be fighting to supply the silicon for one of the fastest-growing segments of computing: smartphones, tiny laptops and tablet-style devices.

The fight pits several big chip companies — each trying to put its own stamp on the same basic design for mobile chips — against Intel, the dominant maker of PC chips, which is using an entirely different design to enter a market segment in which it has a minuscule presence.

Of course, Intel’s (s intc) favorite chip for mobile devices is still the Atom processor, commonly found nestling at the heart of netbooks everywhere. The Atom processor is small, energy-efficient — and terribly slow.

The challenge, then, is clear: make a smaller, ever-more-energy-efficient chip that doesn’t trade performance for low-power-consumption. Steve Jobs, when announcing the iPad to the world in January, hinted that the iPad’s A4 processor might have achieved this lofty goal;

iPad is powered by our own custom silicon. We have an incredible group that does custom silicon at Apple. We have a chip called A4, which is our most advanced chip we’ve ever done that powers the iPad. It’s got the processor, the graphics, the I/O, the memory controller… Everything in this one chip. And it screams.

Mind you, Steve Jobs is the King of Hyperbole, so we should take his claims of speed with a grain (or ten) of salt. It’s encouraging, then, that the feedback from level-headed reporters, and specifically, beloved Mac-head Andy Ihnatko, confirms that, at least when compared to the iPhone 3GS, the iPad is unquestionably nimble;

This thing is FAST. I stretch-zoom a webpage and it keeps up with me now [sic] matter how fast I zoom and scroll. When you turn a page in iBook, it’s not “an animation of a page turning”… you are TURNING a freaking PAGE.

I think, most importantly, this “$1 billion” investment speaks volumes about Apple’s commitment to the iPad and iPhone product families. (We all fully expect the A4 to wind up in an iPhone sooner or later, yes? I mean, that much is obvious, right?)

The iPhone set the stage for mobile touch-based computing and the iPad will soon step into the spotlight. Let’s not forget, also, that Jobs very deliberately (re)defined Apple as a mobile devices company. Apple is taking its touch-based, mobile-computing strategy seriously enough that it’s prepared to spend real money investing in it.

Billions and Billions

Impressively, this isn’t the only billion-dollar investment Apple has made recently. I wrote here back in May 2009 how Apple’s still-under-construction server farm in North Carolina also represents an estimated $1 billion investment. Apple hasn’t confirmed what the server farm will be used for, but it’s sensible to assume Apple is looking to improve and expand its cloud-based services.

You see, a great many of Apple’s mobile devices are going to be connected to the web, so it makes sense that Apple should want to provide end-to-end software and services for its iPhones, iPads and MacBooks. Aside from the obvious aesthetic niceties of Apple-software running on Apple-hardware, the “it just works” ease-of-use of MobileMe and iWork on a MacBook or iPad more than make up for their expense.

Apple knows that if its software works exceptionally well, and is a pleasure to use, people will pay for it despite the existence of free alternatives. Sadly, MobileMe and can only be described as “decent” and “adequate.” Perhaps the server farm is one step toward making them “exceptional?”

However you look at it, the facts speak for themselves; Apple is gearing-up for a future that is focused on mobile computing hardware and services, and its recent sizeable investments and acquisitions are bold steps toward that goal.

$2 billion in two years doesn’t sound like a lot for a company as wealthy as Apple. But make no mistake, it’s still a huge bet, and a particularly brave one, too, given how many tech pundits and punters have failed to understand the utility of the iPad and what it means for the future of computing.

11 Responses to “Apple’s Billions and Billions”

  1. Please look at Apple’s 10Q or 10K. There, they have an item called R&D.

    Yes, I know that the last 10Q only goes up until Xmas, and that the iPad hasn’t yet launched, but when Apple’s Tim Cook and Peter Oppenheimer discuss the last quarter’s earnings on a conference call, they talk about their guidance for the next quarter and year. They typically note any unusual expenses that they foresee.

    SINCE, they made no mention of any unusual expenses to expect in the coming quarter, then one can assume that R&D is growing at its 3% rate it has been. There isn’t any big expenditure on chip development.

    Apple takes an ARM core, A8 or whatever and adds its own extensions. It’s a custom design based around a standard core. It does not cost them a Billion.

  2. Wild Guesser

    People make me laugh!

    “There’s NO WAY Apple spent 1B on this!”

    “500M maybe.”

    Let’s all admit we ain’t got a rats ass idea how much this might cost. Mightve been 300M – mightve been 2B.

  3. Nicolas Doré

    A couple of points: A new chip may cost a billion but the A4 is an ARM core with what seems to be off the shelf graphics logic (it has to be compatible with the iPhone’s hardware, after all). Hardly a “new” chip.

    Also, regarding closed systems, if you’re talking about the chips the PC was a closed system at first (Intel only, after all). ;-) Apple is building closed devices but seems to have learned its lessons as far as pricing and the importance of building an ecosystem around its products. It could also be argued that their devices’ closeness brings benefits to customers (stability, packaging), whereas that was not necessarily the case before.

    Apple have had some good ideas (and bad ones…) and they have been lucky with how unimaginative most other players have been in the MP3 and smartphone markets, but they have also executed extremely well to capitalized on those ideas and lucky breaks.

    • If DOS had been portable across chips, Intel compatibility or “single source” issues would not have mattered as much.

      OS X is portable across processors, and already proven to run on PowerPC, Intel x86 and ARM.

  4. The A4 processor is optimized for mobile applications and battery conservation at the expense of operating OSX. The A4 will no doubt find its home in the next I-phone. Apple once again is also closing its system from hardware to software, you would think they learned the lesson of the past.

    • “The A4 processor is optimized for mobile applications and battery conservation at the expense of operating OSX”

      You do know that the iPhone and soon the iPad run OS X? They don’t run Mac OS X but it is still a version of OS X optimized for mobile devices. Your statement doesn’t make much sense.

  5. I very much doubt Apple spent $1B on the A4, but $500M I could believe.

    PA Semi cost $278M of that, and there’s ongoing expenses since then, but I really don’t see where you could blow $1B on a single chip. It’s not like Apple’s never done this before …