HOW COOL CHIPS WERE DEVELOPED
Isaiah W. Cox, President, Cool Chips plc
Remarks to the Nanotech Planet Conference
San Jose, California, May 14, 2002
I wanted to start by thanking the organizers for the opportunity to speak, and to Dr. Long of Altair, for allowing me to share his time. It is also a privilege to be addressing leaders and pioneers here in the heart of Silicon Valley; what has been accomplished here in the last 20- or-so years has been an inspiration to us, and the whole world.
This is a prototype Cool Chip [show Cool Chip]. It is a little cooling system, which has no moving parts except electrons. They tunnel, using quantum mechanics, to carry heat from one side to another. It's very different from current cooling solutions.
It is said that you need to 'Think Differently' in order to be successful. That is, of course, an equally likely guide to failure. Different isn't necessarily better -- it is just different. But if you want to stand out, you do need to step away from the pack.
The scientific genius behind Cool Chips as thermionic cooling, and many other inventions and patents, is a fellow named Jon Edelson. I met Jon when we were both students at Princeton, and it was clear that he thought differently. Jon figured that the best way to get around on the sheet ice during the New Jersey winter was strap-on roofer's sandals, the kind with the 2 inch steel spikes. He was the only person to wear them. And he was the only person who didn't fall on his butt.
Jon is the only person I know who computed the numbers, including the cost of insurance in New Jersey, and figured it was cheaper to get his pilot's license than his driver's license. And he did. Consequently, he was able to climb into an airplane and fly it -- but someone had to give him a lift to the airstrip.
It didn't occur to Jon at the time that driving has other benefits. And while he could take dates for a romantic midnight tour around Manhattan at under 1000 feet, his pilot's license just wasn't very useful for getting to the grocery store. Thinking differently does open one up to the occasional pitfall.
When we started to work with Jon, when he was still a student at Princeton, it was in large part because his mind would not stop working in very peculiar directions. And he came to work with us because we made business decisions in a peculiarly logical way. We didn't follow a fashion, or think something could not be done because somebody wasn't already doing it. We used *Present Discounted Expected Value*, a big mouthful for something that's really a very simple tool. The entire purpose of Expected Value is making smart decisions when you lack firm information. It comes down to estimating probabilities, which, as any meteorologist can tell you, is a lot easier than actually predicting a sunny weekend.
When Jon first proposed cooling with electrons as a working fluid, back in 1994, we did the math. What were the chances that this guy had really figured out something that everyone had missed? We were new to the field, but it was clear that most companies in the area did not do basic research in cooling -- in fact, they had few or no physicists on staff. The research area was virgin territory. So it suggested that our chances, if we did not violate the laws of physics, were better than zero. They were certainly no better than 5 or 10 percent. We concluded that 1% was a fair estimate. The reward on the other side, however, was immense. At the time, the total cooling market, at over $100 billion yearly, dwarfed the entire semiconductor field. And if we were alone, with a strong patent base, we could potentially reap enormous rewards. The Expected Value was very positive. So we started working in this direction -- filing patents, resolving technical issues, etc.
Thermionic emission occurs when an electron jumps from a surface into a vacuum. It isn't a new idea; its pedigree goes back to the 19th century. Thermionics was once called the "Edison Effect", and it is used in every cathode ray tube in televisions and computer monitors. But nobody had ever used it to cool -- thermionic emission normally only occurs at very high temperatures, while cooling can be done by "opening a window".
Our innovation was to see that thermionics could theoretically be used for cooling, at room temperature and below. We wanted to move towards a commercial device so that we could license Cool Chips to companies that were already in the cooling business. Funding such research is expensive and in 1996 we approached one large refrigeration company that was willing to help support the research -- on one condition. They wanted to see experimental proof of the *concept* of thermionic cooling, that is, cooling across a gap, by demonstrating *any* net cooling. And they gave us a letter of intent to that effect.
So we duly set up an experiment, in a very temperature-stabilized environment, to prove the mere concept of thermionic cooling. And we did it. We proved that thermionic emission cools. Of course, this is known, at least in theory, to any physicist, but the company we were talking to (and many others since) could not paper a funding decision unless there was empirical evidence. So we proved that thermionic emission cools, and we went back to them, and asked for their financial support.
Their reply? "Sorry, you proved cooling, but not *enough* cooling. Keep developing", they said. Our lesson? That we needed to make sure we could pay for our own research, and not expect help along the way.
Our thermionics work progressed, and ran into some materials limits. About this time, we asked one of our technical patent writers, who was based in Denmark, to come meet us in London, to discuss cooling. He arrived in London, and we told him about the thermionic idea. This physicist, Dr. Avto Tavkhelidze, was immensely relieved and intrigued by the concept. He had feared that we were planning to try and miniaturize compressors. Avto said that he had done the math, and concluded that making tiny compressors for cooling was not likely to succeed. When he saw what we *were* doing, he got innovative.
Avto seized on the idea, and proposed that we really needed to look at a variation on the theme -- quantum mechanical thermotunneling.
This is thermionics with a really cool twist -- instead of making the electrons go into a vacuum, just put two surfaces close to each other, and let electrons "tunnel" from one side to the other. A lot of the physics is quite similar to thermionics, and it overcomes the basic roadblocks in thermionics.
It was this development that brought Cool Chips into the world of nanotechnology. For the gap between the two surfaces must be on the order of 2-10 nanometers for maximum effect.
Again we did the Expected Value. Thermotunneling was a smart bet. The risk of failure went down, the reward went up. And we endorsed it. Avto assembled his team, and got to work.
The problem is that while thermionics is hard to explain, thermotunneling, like the quantum mechanics on which it is built, can be impossible to discuss at dinner parties. We had a difficult time conveying it to VCs, and we only tried DARPA once, for which we received a perfunctory rejection. We looked at the totality of government forms and figured it was easier and quicker to do the research than to do the paper work.
Instead, we went it alone, funded only by family and friends and a few visionary investors. And our approaches to industry were, for many years, rebuffed. Why? Because quantum mechanics is so far away from what is comfortably understood.
In 1998, only one year after we first conceived of using quantum tunneling to cool, and long before we had managed to prove the concept, we gave a presentation to a group of scientists and engineers at one of America's largest and most prestigious companies. This company is very involved in the business of cooling, and they had peopled the room with the best and brightest in the field of cooling.
Now this is an old field. Willis Carrier invented the air conditioner 100 years ago. There was still innovation in the 1920s and '30s, but by the end of World War II, all the other ideas had fallen by the wayside, and everyone worked on this one core idea: the compressor, with a refrigerant.
For 50 years all anyone has done in this field has been to work with compressor systems. Innovation was all incremental -- refrigerants were improved, from propane, to freon, to newer chemicals. Motors were improved, making them more efficient and smaller. Manufacturing has been honed, and moved to Mexico. The innovation has been steady, and focused. And all the experts in the field became very specialized at this type of cooling. Along the way, these companies stopped investing in in-house basic research. They shed their scientists, and kept their engineers. At most, they sponsor work at universities, and do the "safe" research -- the stuff government will give them money to do.
Along we come, backed by credible experts, and say we have a new way of cooling which can turn the $100 billion cooling business on its head. So the executives naturally turn to *their* experts to evaluate Cool Chips. The experts, who know a great deal about freon and compressors, but have never been expected to know quantum mechanics, are not always sure what to do. What they want to do, is evaluate is a finished product, a refrigerator in which Cool Chips are a "black box" and they don't need to know anything about the devices themselves. This is an entirely reasonable response to limited information.
Until we have that Black Box, most engineers punt. Unable to evaluate the technology, this particular team wrote a report saying that what we were doing "might be possible", but that "it will take at least 10 years to prove concept".
I relayed this sentiment to Avto, the inventor of the thermotunneling Cool Chip. Avto thinks different, too, though he can express it bluntly. His reply was brief: I think we can do it in three years.
Well, with a budget that was, by comparison with normal corporate research, pennies, we went ahead and got to work. And we proved concept for thermotunneling last year, less than 3 years after we were told that it would take 10, if it could be done at all.
But our work is very far from over. What we are trying to achieve is right out of the nanotechnology playbook: use small things to accomplish great ones. The conventional world isn't sure how to evaluate it. Seriously, cooling hasn't changed for 100 years. Why should it possibly change now?
And, executives know very well how to assess their own risks. Write a million-dollar check for a failure and you wind up running the warehouse in Wichita. The problem is that corporations reward safe behavior and grossly over-penalize risk-taking if it turns out to fail. But Expected Value reminds us of what we already know: if there is no risk, there is no reward.
We have had the same problem with investors (and VCs). Cool Chips as a technology is overwhelmingly superior to compressors, in every measurable way, from size and weight, to power density, efficiency and cost. We know it. We are building it.
But what does a technology company do when someone says, "What is it worth?" We had a simple choice: we could make the number believable, or we could tell the truth, at least as we understood it. And the truth is very difficult to swallow. If Cool Chips are successful, we are aiming squarely at a market which has new sales of over 100 *billion* watts of cooling each and every year. It would take something over 1 billion Cool Chips to service that market, with margins of well over 50%. And to make it sweeter, we'll be profitable after we sell the first 100 chips, for space applications, and continuously profitable from then on.
So we have told the truth. We have told it to our investors and friends since we started this work, and we have told the public. But the truth is very hard to believe. Who are we to make these kinds of claims? We are a small company based in Gibraltar, not in Silicon Valley or Cambridge, with none of the normal trimmings of Big Science.
So most people have concluded that we must be delusional. Nanotechnology, of course, is only now starting to come out of the shadow of delusion.
Operating in the early days of nanotechnology, Cool Chips received no VC funding -- the dot-coms were much more popular. And VCs had the same reservations that major corporations did: it was easy for them to believe that a piece of software could revolutionize a market which did not yet exist, but they found it hard to believe that new industrial technology could make a dent in a mature $100 billion business.
Consequently, we have no competition today, just opportunity. Traditional companies are not the enemy -- they are potential customers. Our business model calls for us build a research chip fab and beyond that to license the Cool Chips technology to the leaders in every market that requires cooling.
We are not planning to do anything that someone else can do better -- we don't intend to make, package, system-design, shock-test, certify, distribute, sell, warranty, or repair products that use Cool Chips. The companies that already make automotive air conditioners and computer chips can do all of these things. They are lean and are world leaders in areas where we cannot (and do not want to) go. We know our limits, and we freely acknowledge that Cool Chips, no matter how important, will always be a component in someone else's system. We can help them make their system better, whether it is a car, an air conditioner, a refrigerator, a cargo container, or an airplane. The dominant players in those fields are likely to remain dominant -- assuming they license Cool Chips.
The opportunity presented by Cool Chips is there for potential licensees and other companies that wish to work with us.
It is also there for our investors. Most smart money defines "success" for a startup venture as the IPO. We do not.
We are basically self funded, which means we have been spending our own money, and that of hundreds of members of our extended corporate family, including shareholders, scientists, patent staff, and others. Spending your own money is a great way to keep the burn rate down. :-) It also means that shareholder value is especially important to us -- we *are* the shareholders.
Since we learned that Cool Chips could not follow the conventional path, we chose to do things a little differently. For us, going public was not an opportunity to cash out. It was just a way to increase visibility, and add some liquidity. So we went public early. We have great lab results today, but no product to sell. Yet.
Going public was again not a case of taking the traditional road. We are in this for the long haul, so what happens now is not as important to us as what happens in a month, a year, a decade.
Rather than spending 8 months and hundreds of thousands of dollars on an Offering Document to float an IPO, we put the effort and funds into our research instead, and we went public through a back door exemption.
It turns out that you can go public in the United States with nothing more than a few hundred shareholders, audited financials, and legal status as a foreign public company. The NASD has to sign off, and they did. We started trading, on no advance notice, just a few weeks ago, on the Pink Sheets. We are now a public company trading in the United States.
For the past 8 years, we have been working in stealth mode, to build our patent library, to advance without competitors. Except under NDA, we have not even breathed the word "thermotunneling", until this conference. We made certain that we waited until we could be sure that the technology was well in hand, that it would be clear to potential licensees that the work remaining between our laboratory prototypes and mass-produced, high-output and high-efficiency Cool Chips, is all engineering, and doable.
So nobody has known about us, until now. This conference is our first opportunity to describe to a professional forum how the technology really works. How we can make a technology with no moving parts that can efficiently do the job of a compressor 20 times its size -- and do it for pennies on the dollar. This is how we think differently, by being not only the first company in a field, but right now the *only* company in a field. In the past 8 years, we have seen the odds of success go from 1%, to where they are now, at about 95%. We see no show stoppers today. We have proved concept, and reduced all the technical issues to issues of engineering. And the reward has skyrocketed as well. We are not just aiming at commercial refrigeration, and air conditioning systems and cars, but also at space-based infrared sensors, and handling the cooling load of the next processor Intel would like to build, but cannot, because there are no cooling solutions out there. That solution is Cool Chips.
The rest looks like plug and chug. We believe that both the science and the economics work, and that in the coming years Cool Chips will come to dominate all the world's significant cooling markets.
And in turn, our job, as management, is to return value to our shareholders. Value, while it can be measured at any moment by the stock price, is fundamentally about long term success, the twinning of a promising present, and a very bright future.
Cool Chips is set to be immensely profitable, and our company is aimed squarely at maximizing that long term reward.
Copyright © 1995 - 2010 Cool Chips plc. All rights reserved.
Forward Looking Statement may be found at http://www.coolchips.gi/fwdlook.shtml.