IBM is on a roll. Their POWER architecture runs IBM's powerful servers and has been adapted to form the basis of the Mac's PowerPC processors. The new Cell processor - jointly developed by Sony, IBM, and Toshiba - will be at the heart of the Playstation 3.

The times they are a changin'.

Microsoft Windows and Intel processors currently dominate the personal computer industry. Linux is a distant second on desktop hardware with far less than one-tenth as many users as Windows, and AMD's family of processors holds a record 16.6% of new PC sales with Intel accounting for almost everything else.

Intel and AMD are competing in the same hardware space, just as Windows and Linux are competing in the same software space. Apple is the odd man out - they make their own hardware and operating system. Macs don't run x86 CPUs or Microsoft operating systems.

Intel x86

Intel's x86 architecture, which traces its roots to the Intel 8080 CPU of 1974, absolutely dominates the desktop, laptop, and server markets with well over 90% of the market - but I believe in a PowerPC future.

The x86 architecture has become incredibly powerful and complex, and that complexity may be part of the reason Intel has been unable to build a 4 GHz Pentium-class CPU. Intel has tweaked their design to get more and more performance out of 3.x GHz processors, but speed increases have been slow in coming in recent years.

Intel's CPUs are also power hogs, and that's why Palms and Pocket PCs don't use x86 processors. The Xscale processors found in most Pocket PCs comes from Intel, but it uses ARM architecture, which is optimized for low power consumption and long battery life.

Power Inside

Even Microsoft seems to realize that x86 is running out of steam. Their next generation Xbox won't be based on Intel components - it will run a customized IBM Power processor. Nintendo's Game Cube already uses an IBM PowerPC processor, and Sony will be adopting the Cell processor (a Sony/IBM/Toshiba project that's also part of the PowerPC family) in the Playstation 3.

Imagine that - the top three gaming systems will all be using the IBM Power architecture next year. And last week IBM announced an initiative to promote the Cell architecture beyond gaming.

It's easy to imagine places where the Cell might be an ideal solution. TiVo and other digital video recorders don't need huge amounts of processing power, but they do need good video processing. The Cell could be a perfect fit.

The Cell might be a good choice for networked storage devices, since it draws little power and manages multiple tasks simultaneously.

Another place the Cell processor might appear is the "smart" terminal market, which is currently dominated by Wintel PCs. A smart terminal needs to have enough memory and power to do simple tasks - log into a central server in a retail store or display Web pages and email in a library.

IBM has set their sights a bit higher than that, though. They want to become a crucial player in the high-end imaging market, which includes aerospace, defense, industry, and medicine.

As many observers have commented, the Cell is cousin to the PowerPC processors found in Macs since 1994, and it should be feasible to build a whole new personal computer architecture around the Cell processor.

OS X Everywhere

A new architecture needs an operating system. IBM has been supporting Linux develpment heavily for years, and it's one option, although it falls far behind Windows and the Mac OS when it comes to user friendliness. (This is the one area Linux geeks need to address if they ever want to compete in the broader consumer market.)

Microsoft hasn't made Windows for PowerPC since the Windows NT era, and that was a fiasco, so they may be wary of developing Windows for a different hardware platform. That would divide their focus and confuse the market. It could also cut into Microsoft's bread and butter sales to companies backing the x86 architecture if the Cell really takes off.

The third alternative is Mac OS X, which already runs on the PowerPC family of processors, has the stability of Unix and Linux, has the user friendliness of Windows, and has no known viruses or other types of malware.

Is Apple a Hardware Company?

The last time Apple tried to expand the Mac OS base by licensing clones, it nearly destroyed the company. Instead of seeing Maclones marketed to new computer buyers, they were mostly sold as new computers to current Mac users. And they almost always offered more power for less money than Apple. Apple hardware sales took a beating, and license fees weren't enough to offset those losses.

Apple's attempt to redefine themselves as an operating system company failed, and since then the company has been seen primarily as a hardware company. Without the Macintosh hardware sales, there is nobody to sell OS X to.

But the software company vs. harware company comparison is rooted in the Wintel market, where almost every computer runs Windows and has Intel inside. Apple hardware and the Mac OS provides a unified platform - but it doesn't have to remain that way.

Let's say companies such as Dell, HP, Gateway, and the like are tired of Microsoft's hardball tactics and Intel's anti-AMD policies. They want to make money selling computers, don't have the resources to create an operating system, and want something that would differentiate them from the zillion brands of Wintel PCs currently on the market.

Now imagine that Apple and IBM come together to design a hardware reference platform that anyone can license at a reasonable price, one based on the Cell processor. The computer could ship with Linux, which costs the computer makers nothing.

Apple could work with manufacturers to either license OS X for factory installation or include a coupon with every new Cell PC offering a special price on Mac OS X and the iLife suite.

It's not the same thing as cloning. The Cell PCs are perfectly usable with Linux, manufacturers are no longer beholden to Intel and Microsoft, buyers save money, and OS X becomes the option of choice for those who want something easier and friendlier than Linux.

We all win, although sales of Apple hardware might decline. Or maybe not. With the Mac mini, Apple showed they can produce innovative, marketable computers with lucrative pricing.

The new platform would compete with Macs, but it would also create a visible alternative to Wintel, creating a platform that could slowly eat away at the near monopoly power of Microsoft and Intel.

If IBM and Apple pursue this option, the personal computer market could be a very different place five years from now with Power Inside.