Judging from recent comments by Chenming Hu, he's devoting more study to the fable-something others in the industry ought to consider.

Earlier this month, at Taiwan Semiconductor Manufacturing Co.'s San Jose technology forum, the popular technologist bucked his own company's trend setting by saying that the industry ought to consider a slower march toward tinier circuitry. The reason is that end users aren't sucking up the advantages like they used to.

In other words, the industry racing along like a hare means little if end users are moving like tortoises. And we all know they are.

But bragging rights are a prized measure of self-worth in this industry. Let the process guys forge ahead, right? And leave it up to the application engineers and marketers to find nifty uses for their gee-whiz accomplishments. Sounds good. But if that were always the case, Nokia wouldn't be looking for growth by selling diamond-studded phones.

What Hu says is worth some serious consideration, but I doubt anyone will buy into it. Even Hu knows TSMC will likely continue its scaling race and reap the bragging rights attendant to outpacing the International Technology Roadmap for Semiconductors.

Nevertheless, he suggests going back to introducing new technology nodes every three years instead of the more recent development of every two years.

From 1965 to 1995, the industry followed a three-year cycle like clockwork. During that time, sales grew an average of 15 percent each year. Now take notice of the 1996 to 2001 period, when the industry accelerated to a two-year cycle. Last year, sales rang in at about 1995 levels and profits were lower, Hu said, making it the worst six-year period on record.

Scaling "improves the performance-cost ratio and inspires new applications in the market," Hu admitted. But it's not enough. Toward the end of that six-year period, the industry started to bellyache about the lack of a "killer app" that would sustain the party, driven by a need for those 2-GHz Intel toys brought about by the wondrous benefits of scaling.

"Markets . . . need time to digest the benefits of technology scaling," Hu said. "Perhaps we will go back to three years every generation. That, I think, would be a good thing rather than a bad thing."

And it would give all those hard-working, tech-shy turtles (aka consumers) a chance to catch the hare.