I was talking to a new coworker today and of course the topic of our first PCs came up. It was Cyrix-based. I didn't mention my first PC (it seems I'm about four years older--it was an Am486SX2/66).

With only a couple of exceptions, I've always bought non-Intel PCs. Most of the Intel PCs I have bought have been used. One boss once went so far as to call me anti-corporate.

I'm not so much anti-corporate as I am pro-competition.

I was a second-generation AMD fanboy, not first. When the Am386DX/40 hit town, I was aware it was the best value in the industry, giving better performance than an Intel 486SX/25 for the price of an Intel 386DX/33, but I didn't really care because I was still an Amiga guy at that point in time.

But that's the reason I'm an AMD guy today. One of the reasons the computer market is so stagnant today is because it's dominated by Microsoft. There's nothing exciting going on there. There hasn't been anything exciting coming out of Microsoft since the mid 1990s when they had to compete with OS/2. OS/2 never captured a huge amount of market share, but OS/2 promised to be a better DOS than DOS and a better Windows than Windows, and to a large extent it delivered. Say what you want about OS/2, but I could load Tony La Russa Baseball 2 up in a DOS Window under OS/2 and it would run faster than it ran under DOS, even though the game didn't have the machine's full attention. OS/2 2.1 (and later 3.0) had Microsoft running scared, because it ran all of the software that was available in the early 1990s and it ran it quickly, in a fully pre-emptive multitasking environment. Microsoft responded with Windows 95 and Windows NT 4.0 because it had to--Windows 3.1 just couldn't compete with OS/2's stability, and booting into DOS using a custom boot disk for every game they wanted to run wasn't something the public was going to put up with forever.

And what's happened since then? Windows 98 was basically a service pack, improving the stability of Windows 95 but not offering anything revolutionary. Windows 2000 was a lot better than NT 4.0 but not the kind of jump that Windows 95 was over 3.1. Windows XP did a lot to improve backward compatibility with old DOS and Windows 9x games, and while it was a big leap from Windows 98 or ME, it wasn't a tremendous improvement over Windows 2000. I haven't heard anyone say anything good about Vista. At least the Windows 95 box was pretty, but Vista doesn't really even have that going for it.

Microsoft's primary competition today is illegal copies of its own operating system, so its main concern with Vista is keeping people from making copies of it. And it shows.

Apple is trying to compete, but its market share is around 10 percent. We don't exactly have a duopoly.

When I got interested in computers in the 1980s, there was all sorts of interesting stuff going on. IBM and DOS were things you used at work to do accounting. At home we used all these weird and wonderful 8-bit computers that were technically obsolete, but engineers kept figuring out how to squeeze more capability out of them. Revisionist historians talk about Apple dominating the 8-bit era, but that wasn't true. At its peak, Commodore sold as many C-64s in a single year as Apple sold Apple IIs in that line's entire lifetime. Although Commodore was the king of sales, Atari arguably had the best 8-bit computer (the 800/XE/XL family). Tandy had its Color Computer line, and while it couldn't match the graphics and sound capability that Commodore and Atari had, it had a far more powerful CPU. Coleco's Adam is little more than the butt of a joke today, but on paper it should have done well. Coleco took several chips that Texas Instruments had used in its failed TI-99/4A, paired them up with a more conventional CPU (the popular Zilog Z-80), and made a competitive computer with it. Its biggest problem was that it was late to market and plagued with reliability problems at first. Kind of like Windows Vista.

And that's the beauty of competition. In the 1980s, if you delivered a product like the Coleco Adam, you went out of business. But if your name is Microsoft and you have 85% of the marketplace, you can deliver something like the Coleco Adam and keep on chugging.

The really exciting stuff in the 1980s wasn't in the 8-bit arena though. The Motorola 68000-based computers were where the action was. The most famous of these, of course, was the first-generation Macintosh. But the Atari ST and Commodore Amiga used the 68000 too, and unlike the first Macs, they paired the powerful CPU with color and powerful sound. The three companies threw bricks at each other a lot, but they kept each other honest. Apple ended up having to add color and sound and expansion slots to its Macs in order to compete. Commodore designed a low-cost Amiga to compete with Atari, and a higher-priced model with lots of drive bays and expansion slots to compete with Apple.

These three companies, ironically, built the kind of machine Bill Gates tried to get IBM to build in 1981. Gates wanted IBM to use a Motorola 68000, and then they would have used Xenix, Microsoft's version of Unix, for an operating system.

The result was Microsoft trying to play catch-up. Windows was in development before these machines hit the market, but Microsoft knew the Mac was coming long before it happened. Microsoft had a prototype and was one of the first Mac developers. In typical Microsoft fashion, Microsoft was talking about Windows in 1983, but didn't deliver anything until 1985, and what they delivered wasn't useful for very much. It wasn't until Windows 3.0 came out in 1990 that it hit prime time. By then the code was stable enough that you could use it for a few hours at a time, and PC CPUs were powerful enough that Windows could keep up with an Amiga or ST or Mac without embarrassing itself.

And that was the beginning of the end. It was one thing for Commodore and Atari to compete with 286 clones that could barely run Windows. But within a year or two, they were competing with Tandy 386s that sold for $1,200 at every Radio Shack in the country. Whether you lived in New York City or Buffalo, Missouri, you could walk into Radio Shack and see a computer running Windows and buy it on the spot. Neither Commodore nor Atari had a dealer network anything like that. And if you lived in a big enough city, you could walk into a "superstore" like Best Buy or Circuit City or Silo that were sweeping the nation at the time and buy a Packard Bell for even less.

By 1993, Commodore and Atari were non-factors in the marketplace. It was down to Apple and the PC clones running Windows.

So what does any of this have to do with AMD and Intel?

AMD and Intel keep each other honest. When Intel released the Itanium, AMD countered with its AMD64 architecture. While the Itanium gives better 64-bit performance, AMD64 does a much better job of running the 32-bit applications we all run today. Itanium has a better long-term approach, but it's designed for a future that will never come on its own because people still want to play their old copy of The Sims and have it run well on their new computer. AMD's answer for 64-bit computing was AMD64, and its success forced Intel to clone it.

The CPU isn't as important today because Intel and AMD are making CPUs that have more power than today's software knows what to do with. But that's not Intel's fault, and it's not AMD's fault. Microsoft can't think of a good use for all the power or a way to harness it, and the industry doesn't have the scrappy underdog companies like Commodore or Atari anymore to figure out a use for them and drive the industry.

But if one company had a total monopoly on CPUs, I'm afraid of what I'd see. Probably Intel would become more like Microsoft, delivering products that ran slower and at a higher price than each previous generation. It's unnatural, but it's the norm for a monopoly.

I've heard myself saying several times over the last three or four years that I don't like computers anymore. But that's not exactly true. Either I don't like modern computers, or what HP and Dell sell today aren't computers.

If either Intel or AMD were to succeed in squeezing the other company out of business, the modern computer would become even more underachieving and uninteresting than it is now.