Friday, September 17, 2004

Revolutions happen to the unprepared


I'm glad to see CNET News pick up on the nanotech revolution that's been forced on the semiconductor industry. Executive Editor Charles Cooper, in a column titled Catching the nano wave, shows a deep understanding of why nanotech is not just a trendy word to chipmakers, but something they need to learn or die:

    As the end of the complementary metal oxide semiconductor, or CMOS, life cycle approaches, chipmakers find themselves in a race to catch the wave of the next big technology trend. Consensus opinion has it that we're talking about nanotechnology.

    That's why the Semiconductor Industry Association (SIA), a trade consortium representing American chipmakers, wants to create a Nanoelectronics Research Initiative. If the plan comes alive, a select number of universities will become test grounds for the development of advanced nanotechnologies.

    "We can't even think about not doing this," said George Scalise, a veteran chip hand who now heads the trade group. "We have to make it happen." More here

Indeed, the SIA expects current lithographic methods to reach their absolute limit in 15 years. By that time, it'll be like a Midtown Manhattan subway during rush hour on every chip. You just can't pack them in there any tighter.

So, the group is asking the U.S. government to chip in more money and it's proposing a research institute that will discover what comes next. The goal? Creating an entirely new industry, with new switches, interconnects, materials, memory and manufacturing methods by 2020.

Remember, that 2020 date is the absolute drop-dead deadline. Companies not only need to learn and implement new methods before then, they also have to do it in tandem with evolving standards for new nanolithographic and nano-assembly techniques. So, the various processes and materials that nanotech researchers are experimenting with take on a whole new sense of importance. The semiconductor industry no longer views them as a curiosity. It's thinking 15 years down the road and wondering how the "breakthrough of the week" can scale up into large-scale manufacturing and become incorporated into evolving standards.

Now, contrast that to the strategy of some of the top nanotechnology companies with their eyes on the semiconductor market – ZettaCore, Nanosys, Nantero. They're all focusing on building technologies that can be incorporated into today's semiconductor processes. They're doing this because that is how they can obtain VC funding, impress investors and stay alive in the short term. But in doing that, they're essentially applying some chewing gum and nails to hold together a house that belonged to their parents, when what they could really be focusing on is preparing for the day the whole rotting, mold- and termite-infested edifice simply collapses under its own weight.

With VCs overcompensating for their reckless enthusiasm during the '90s, they're missing out on an opportunity to really "change the paradigm." We're not talking about a couple of college kids with a bar-napkin plan to sell kumquats online. We're talking about coming up with the standards for nano-electronics that will drive the industry deep into this century.

So, here is where the ZettaCores and the Nanteros need to lead, not follow. Unfortunately, they see themselves as merely startups beginning from a position of weakness, begging for crumbs from the big dogs. They need an attitude change if they're interested in truly enabling the next revolution. The nanocompanies with the right IP and the right timing can pounce and grab, leaving all the "incrementalists" and "integrationists" scratching their heads and wondering how and why they were left in the dust.

In the late '70s, my brother, now a veteran of many tech startups, was tinkering around with the demo model of Radio Shack's TRS-80 at our local shopping mall. It was there that he wrote his first program – a computer sketch program called the Doodle Pad – and released his first product while he was still in high school.

Sometime in the early '80s, though, I overheard him talking to my father about IBM's plans to introduce a personal computer for the consumer market. My big bro correctly predicted that Radio Shack and Apple were about to be blown out of the water. It didn't take a genius. IBM was, and still is, well … IBM. They have a way of dominating any room they enter.

But the reason my brother is not a multizillionairre today is that he had no way of predicting what was not obvious: Before 1981, no analyst would have told you that a couple of greasy-looking kids who had just moved their startup from Albuquerque to suburban Seattle were about to take over the world. Digital Research should have had the goods to deliver IBM an operating system for its new personal computer. It didn't. Bill Gates did. Historians can argue the hows and whys, but for our purposes that's the end of the story.

There needs to be a recognition that when revolutions happen, they usually happen to the unprepared.

NanoBot Backgrounder
Welcome to our Nano Nightmare
Thanks for the NanoMemories
'Terabyte territory'

Related News
Intel's Frank Robertson to Keynote SEMI NanoForum (nanotechwire.com)

Brain power will win nanotech wars: Semiconductor executives expect formidable developments at college (The Albany Business Review)

2 comments:

Anita Campbell said...

Nice article, Howard.

It's always a real challenge for young companies, figuring out how to jump on the future opportunity of tomorrow but at the same time paying the rent and electric bill today. In some ways it is better to be "two guys in a garage" instead of a young company, because once you start getting a little money coming in and hiring people and acting like a real business, it's gets harder and harder to take forward-looking risks. Companies become more risk averse, because they have more to lose at that point.

Anonymous said...

Howard,

As Richard Feynman said, "For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled."

Zettacore is pushing along. They truly have working arrays. Lots more work to be done, and lots of issues with the market - the technology is a DRAM replacement, not universal memory yet. And one need only look at the DRAM market to realize the Zettacore has an issue going public anytime soon. However, Zettacore's technology is real and is a legit technology that is scaleable into manufacturing.

Nantero's NRAM is Feynman's poster child of "cargo cult science." Nantero is trying to fool Nature. Nature is winning.