Hand-Me-Downs

T he best technology never wins--or so goes the old maxim. The original IBM PC is often held up as an example of the maxim's validity. Why, for example, did not the Amiga become the PC of choice? (Kindly disregard any unfortunate resemblance this discourse may have to an old religious war.) I first encountered the Amiga computer within the pages of the premier issue of Amiga World Magazine back in 1985. At that time, it was the most exciting personal computer in the world. Based on the 68000, the Amiga had built-in, high-resolution color graphics and stereo sound, a preemptive multitasking operating system, a graphical user interface, and a command line interface with a lot of the features and power of the Unix shell, all for a little over a thousand bucks. The Amiga also had its share of drawbacks, such as blinding flicker in high-resolution video mode (because of NTSC compatibility) and icons that were way too cute for Wall Street. The drawbacks, along with the worst marketing since the New Coke rollout, conspired to drive the Amiga to virtual extinction.

At the time the Amiga was introduced, the Mac had not found its desktop publishing niche, and the IBM PC had not yet become ubiquitous, so theoretically the Amiga did have a market opportunity. No one will admit it, but I believe that it may have spurred improvements to the PC and Mac. Ultimately, the Amiga was unsuccessful, at least in terms of its original target market, although it did find a niche as the platform for a lot of video development. One of its most popular applications was as the engine for the video toaster, a video effects system from Newtek. Rumor has it that at a MacWorld conference, a Mac was exhibited with a video toaster using an Amiga as the front end.

Many good architectures fail to find a place on the desktop, a misfortune that can ultimately benefit embedded systems developers. Here's what I mean: with the PC motherboard market essentially sewn up, hardware vendors must reach the desktop via popular add-in boards or--better still--occupying a socket on the motherboard. Getting on the PC motherboard is tantamount to dying and going to heaven. There are only a limited number of strategies to achieving this kind of success. One is to develop a processor that will become the CPU of a successful line of computers. Not something I'd want to invest my money in. Another strategy is to develop a coprocessor that will be incorporated into most systems. For example, let's say someone develops a new, whiz-bang multimedia chip with an eye on getting socketed onto the PC motherboard. As PCs evolve into multimedia systems, opportunities arise for this to happen. What happens if for whatever reason the chip fails to gain a foothold on the motherboard? A lot of money has gone into the investment.

As hardware developers look for markets to help recoup their investments, embedded applications begin to look attractive. For embedded developers, this represents a windfall. Processing power often comes with a lower price tag in the embedded market than on the desktop. That whiz-bang multimedia chip may very well find its way into embedded applications. Maybe the best technology wins after all--as long as it's embedded.