Working Up

Here is an update on Michael Arrington’s attempt to build a simple, inexpensive, web-browing tablet computer. One issue I see is that he is learning that it may cost $299 to build this “simple” computer. We can buy fully-functional portable computers for that price. Once again, the general purpose computer makers have bypassed someone trying to make a special-purpose computer. I have seen this happen again and again for over 25 years.

Back in the 1980s (please stay with me as I won’t tell you how I walked through snow in Louisiana to go to school), I was researching  computing architectures for image processing. I saw this same pattern then.

1. People drew the optimum architecture for a computer to perform image processing (e.g. PhotoShop) operations.
2. These people then went about building these computers.
3. The optimum architectures were realized.
4. Problem: by the time the paper design became a functioning computer, you could buy a computer than performed the image processing much faster.
5. The from-the-store computer cost about 1/1000th the amount of the optimum architecture.

Those darned commercial processor makers like Intel, AMD, Motorola, and others just wouldn’t cooperate with the optimum architecture architects.

Patience with the commercial processor makers always won. The story doesn’t seem to have changed any today. Those commercial companies have lots of money and lots of smart people working around the clock. It is tough to beat that.

Lesson (sometimes) Learned: Be patient. Wait for the commercial machine to do what you want to do.