This weekend Intel reported that it had made super-tiny transistors that will power the processors of the future. More importantly, it made them using conventional processes, implying that the development curve will continue at a steady pace. That's good news for the entire semiconductor industry. As the products improve, demand for them will naturally follow.

	Email this page Format for printing Reuse/Reprint
	Related Links Intel's Outlook Improving
	Discussion Boards Intel

By Brian Lund (TMF Tardior)
June 11, 2001

Intel (Nasdaq: INTC), contrary to the predictions of popular pundits, refuses to stagnate and drift into irrelevancy. It continues to develop the essential technical core of future computing and communication devices. The latest offering came this weekend, when the company announced that it had developed tiny electrical switches for future semiconductor products.

Did I say that the switches were tiny? I should have said tiny tiny. Intel has developed silicon transistors 70 to 80 atoms wide and 3 atoms thick, capable of switching on and off 1.5 trillion times a second, making them about 1,000 times faster than those in the current generation of microprocessors. What makes Intel's transistors so important, though, is not their small size -- others have made similarly small transistors in the past. The key development is that Intel made them using the same equipment that it uses for today's transistors.

The implication is that Intel doesn't need to make another technological leap to arrive at a microprocessor that uses the nanotransistors, which probably won't happen until about 2007. The transition to faster processors can follow a continuous path that doesn't require a radical shift in production and capital spending costs.

In a sense, that's bad news for Intel. It means that there is probably not a prohibitive intellectual-property barrier to faster transistors that Intel could erect in front of competitors like Advanced Micro Devices (NYSE: AMD). Instead, this is an advancement that bodes well for the entire industry.

The ridiculously named "Moore's Law" refers to the 1965 prediction that processing power would double every 18-24 months. Obviously this is not a physical law (and has no business being called one), but an estimate for the near future that has proved quite accurate. It wasn't that long ago that even Dr. Gordon Moore, author of Moore's Law, questioned the likelihood that it would continue to apply. He thought that physical limitations would inhibit development beyond 0.25 microns (millionths of a meter).

Well, we're on 0.13 microns now, and Intel is saying that we can get to 0.045 microns. Development continues, applications build upon them, and businesses want to exploit them. Sure, we're in a capital spending drought right now, but it will not continue forever. Indeed, both Intel and AMD said last week that they expect the personal computer market to stabilize in the second half of the year. Long-term, both companies will remain highly relevant. Don't let the current downturn blacken your vision of the future.

Brian Lund aspires to having a candy bar named after him. Wouldn't you buy a Lunderbar? To see his stock holdings, view his profile. The Motley Fool is investors writing for investors.

Feedback about News & Commentary? Please send mail to news@fool.com.