Navigation:
« Cory Doctorow a Liability to EFF | Main | Gnomedex Wrapup »
June 23, 2005
Intel Inside...Apple?
A number of people have recently asked me what I think about the Apple/Intel deal. This has been flogged to death on the net already, so I’ll just add a few brief observations.
Apple is beginning a 2-year transition from IBM’s PowerPC chip architecture to some undisclosed Intel chip architecture. Most everyone is presuming that the Intel chip will be some current or future version of the Pentium 4 and Pentium M, and that seems like a good supposition. But I also wouldn’t be surprised if there were some other chip that Intel has under wraps that might be the subject of this deal.
Apple will likely suffer a mild case of the “Osborne Effect” whereby it loses sales while customers stop purchasing the existing products in favor of waiting for the new ones, but I don’t think that it will be too severe. After it blows over, Apple’s sales will return to able where they are now. Switching to Intel CPU will not boost Apple’s sales by any significant about. There’s no advantage for a Windows user to buy a premium-priced system from Apple.
Intel will be able to increase it’s sales by a few percentage points without having to take them away from AMD. In other words, the market for x86 chip expands with 100% of that expansion going to Intel.
Apple will design their Intel-based systems so that they are architecturally distinct from “IBM-compatible” PCs, and Apple will make sure that the MacOS only runs on genuine Apple hardware. People tend to forget that in the early 80’s there were several different flavors of 68000-based workstations around that wouldn’t run Mac software (Sun, Apollo, etc.), and there were x86-based systems that weren’t PC-compatible (DEC Rainbow, to name one), so this isn’t so hard to do.
However, Microsoft (or some third-party) will make it possible to run Windows natively on Apple hardware. This MAY make Apple hardware more acceptable in a corporate environment due to the fact that it will be theoretically possible for that hardware to run Windows. But corporate standards tend to be pretty exact, so I’m incline to doubt it.
The more obvious way of running Windows and Windows applications on Apple hardware will be through emulators such as Virtual PC. I would be shocked if these emulators weren’t updated to run Windows apps at nearly full speed on a Mac.
Due to the ubiquity of Windows applications, a Windows emulator such as Virtual PC will become almost standard equipment on an Intel-based Mac. This should result in a small increase in Microsoft’s Windows licencing revenue.
In the longer term, this deal MAY result in fewer Mac-specific applications being built. Today’s applications are written to operating system API’s, not CPU instruction sets. Developers who are already committed to the MacOS API will likely continue to development Mac-specific applications. But other developers will likely rely on the emulators to get their Windows API applications running on the Mac. As the cost of maintaining two separate code bases—one with a very large market share and the other with a very small one—continues, developers may decide to abandon the Mac API and concentrate on the Windows API exclusively, relying on emulation to cover the Mac users. This is especially likely if the emulators can be improved to the point where they can run applications without having to reveal an entire Windows operating environment. (If that last bit isn’t clear, send me an email at “scott at trotternet dot com” and I’ll try to clarify.
So, to summerize:
Apple: Short-term loss of sales due to Osborne Effect, but recovering to current market share levels. In and of itself, the CPU switch probably won’t entice many current Windows users to switch to Mac.
Intel: Sales increase of 2-3% without having to battle AMD.
Microsoft: Windows licence revenue increase due to increased use of emulators on Macs.
Windows developers: Slightly larger potential market for their products, but probably not enough to get them really excited.
Macintosh developers: May eventually abandon native Mac applications in favor of relying on Windows emulators.
Now, all this is based on the participants stated intentions at this point in time. The deal also opens up some interesting possibilities should the parties—mainly Apple—choose to take advantage of them.
In spite of their stated intention to not allow the MacOS to run on non-Apple hardware, they could easily change their minds further down the line. This would effectively kill their hardware business and transform them into a pure software company much like Microsoft. I don’t think that this will happen as long as Steve Jobs is in charge. After all, one of Steve’s first actions upon resuming control of Apple was to kill off the Mac clone market.
Apple might try to revive their “Switch” campaign by offering a limited version of the MacOS which would run completely from a CD-ROM in much the same way as a version of Linux does not, thereby allowing existing Windows users to “test-drive” the MacOS on their own systems before they (hopefully, from Apple’s POV) buy a Mac.
That’s all for now. Drop me a line and let me know what you think.
Posted by Scott Trotter at June 23, 2005 11:47 PM