Tags:
Apple,
Intel,
IBM,
MarketingI've been reading the many opinions being bandied about on this Apple Intel deal. A lot of people are just looking at the technology pieces of this Apple/Intel deal and trying to see how those fit. But an awful lot of opinion currently looks like this:
- An unexplained business miracle happens that converts Steve Jobs to the dark side.
- Apple goes Intel x86 for Macs.
- Another business miracle happens that allows Apple to negotiate this transition well.
- Apple profits, goes belly up, becomes Microsoft, or just ticks off its entire customer base
But the truth is this is a business deal, one in which both Apple and Intel want something. We know Jobs wants better parts supply and would love better prices and volumes. But what does Intel want from a deal like this? Why do they care? After all, Apple probably wouldn't make it into the top five OEMs for Intel, so why would they want them as a customer this much? And why would Apple start rolling this out with mobile devices, as claimed by the Wall Street Journal?
The answer? Intels need a lead vendor to launch a Centrino-like platform with WiMax wireless.
Intel has invested tens if not hundreds of millions in this wide area wireless technology. It views WiMax as the obvious upgrade for both individuals and businesses. The only problem: no one is stepping up to build it in as a standard part of every mobile computer. There's a chicken-and-egg problem here: if no computers have WiMax, no router manufacturers or carriers have any interest in it either. Intel needs a lead user.
Oh, one more thing. Probably the most visionary and highest profile executive in Intel below CEO Paul Otellini is Executive VP Pat Gelsinger. He just got moved to head up the platforms business. So any Intel deal with Apple probably has a platform component to it; it can't be just a part sale.
So here's the deal that I believe Jobs signed up for. I believe that Intel signed up to make low-power Intel G5 processors (probably dual-core ones) for laptops. These G5 processors are a little different from ones to date, because they talk to standard Intel platform components (e.g., Northbridge and Southbridge chips and the like); heck, they might even design it to use the same pinouts as a Pentium M. But more importantly, they also get bundled with a WiMax wireless chip. This is the part that they plan for 2006. They then plan another high performance part for 2007.
But because Intel is signing up for a lot of design work, it wants a concession from Steve. It wants the platform design to be licensable to other OEMs like Dell and Acer. After all, if only Apple makes it, it isn't a platform -- it's a proprietary design.
Apple would like this approach instead of a typical Intel x86 approach because it eliminates "the Osborne effect"; that's the case where your customers stop buying your current products as soon as you announce your new one. And given these products won't arrive until 2006, the Osborne effect would be fatal. So Intel would have to concede producing G5 products, but in turn, Apple would have to concede the platform idea.
This is the deal that I think is the most likely scenario past just using WiMax parts. Why? Because the deal gives Apple a stream of highly competitive products (which hasn't always been true), while also providing a key differentiation: the ability to use wide area WiMax wireless out of the box. And Steve doesn't mind reopening the Mac clone business because he now runs a very successful software company, something that wasn't true when he took over Apple.
Think about it. Let's imagine that Dell decides to build a Dell-branded Mac. What OS does it run on those boxes? They probably don't run Windows, because there is a ton of intellectual property in Mac machines that isn't part of the processor and needs software support. And Microsoft is busy during this time frame with Longhorn. So Dell licenses Tiger. How much does profit does Apple make on additional Tiger sales? Lots (think Microsoft profit margins). How about iWork and iLife? Shake? Final Cut Pro? Same answer.
So the bottom line: this could be a tipping point for Apple. Monday could mark the day that Apple begins the transition from proprietary Macs to Mac platforms. But Steve is doing the deal for marketing reasons, not technology ones. And if he has played his cards the way I think he has, Apple stands to make a ton of money in the process, which, after all, is what marketing is about.
P.S. Monday update: I've blogged this separately, but a lot of people are just being directed to this post, so I'll append it here.
A few people have asked why Intel would break with its X86 tradition and make Power processors for Apple? I have three good reasons:
- Sony
- Nintendo
- Microsoft
With all these consoles using Power chips, and gaming console volumes in the millions of units per year, Intel wants a piece of that business, and they won't get it without making Power chips. An Intel Power chip would be good business for Intel and would provide risk reduction for these three big names at the same time so they don't end up over a barrel with IBM in the future.