Moore looks back and forward
During the celebration, the event featured the premier of an excellent video about Moore's law that includes a recent interview with Moore himself on its origins and its future. For starters, Moore reminds us that the electronics industry is built on chemistry -- Moore himself had a PhD in chemistry -- and explains how innovations in material formation made the semiconductor possible.
Moore himself has been amazed by how far innovation has driven his predictions. There has been growth not just in transistor density, but in many other dimensions of the semiconductor manufacturing process as well. For example, when he first made the forecast, a 2-inch silicon wafer was aggressive, while we now have 12-inch wafers and are looking towards 18-inch versions. The result has been a nearly 1 billion times decrease in the cost of an individual transistor. Moore points out that the number of transistors in existence has long since surpassed the number of ants on the planet, and that the transistor is the highest-volume product in the history of mankind.
When asked about the future of the traditional interpretation of Moore's Law of exponential progress in semiconductors, Moore says it's drawing to a close. He cites the atomic-scale and speed-of-light limits that silicon chips are approaching as good reasons to expect a slowdown this decade. He closed on an upbeat note, though. Since exponential growth has become so important to innovation, he explained that, "If Al Gore invented the Internet, then I invented the exponential." You can see the entire video online, courtesy of the Chemical Heritage Foundation:
To ruminate on both the impact of Moore's Law and forecast its future, the event featured a pair of valley pioneers who were instrumental in its early years. Bill Davidow, an early Intel executive and successful venture capitalist, and Carver Mead, a Caltech professor who was a pioneer in semiconductors and an early colleague of and inspiration for Moore, were interviewed by David Brock -- co-author of Gordon Moore's biography. While some of what they said reiterated what is known, they shared a number of new insights.Bill Davidow explains how everything is going to change
Davidow started his remarks by looking back at the importance of order-of-magnitude changes in technologies, such as the horse-drawn transportation making city-states possible, and the railroad making industrial cities possible. Each of these innovations occurred over the course of centuries. He then compared that rate of progress to the nearly eight orders of magnitude we have achieved in processing power and communication speeds over the decades since the introduction of the integrated circuit. He linked that progress to a fundamental shift in how fast and inexpensively information now travels, especially compared with the speed at which people can travel.
For example, he cited that the conventional model of shopping -- such Walmart -- required people to move to the information, calling the store a "file cabinet full of data about what products are available and at what quality and price." Now, that same information moves to the consumer, who can shop online from home and have the product whisked to them. He made a similar comparison about workplaces. It used to be most practical for employees to all commute to office buildings. Now, the Internet makes it possible for employees to work at home, or anywhere in the world.
When asked by moderator David Brock, author of the new biography of Gordon Moore, what that meant for our economy going forward, Davidow presented the somewhat radical notion that it would result in a construction boom. He pointed out that almost all of our infrastructure has been built under the old model of moving people to information, and it will need to be rebuilt as we understand the implications of data being moved to people. He also strongly suggested that we need to rethink our economic models. Innovation is making many things less expensive. Economists measure that as meaning deflation, which they believe is to be avoided. However, for the consumer, lower prices create an improvement of quality of life. So even our economic models need to be revised to incorporate the rapid pace of innovation. Mead echoed these observations, and suggested that many of the problems are addressed in George Gilder's new book, Knowledge and Power.
Carver Mead explains the true significance of Moore's Law
There are few deeper thinkers than Carver Mead. Many of the observations Moore used in making his initial predictions came from Mead. He was an early evangelist for the transistor and integrated circuit. His stories of how hard it was to convince people that first the transistor, and then the integrated circuit, represented the future are a reminder that technology changes always look more obvious in hindsight. In particular, he said that all the research had been going into the transistors themselves, and that his fundamental insight at the time was that the interconnects were going to be even more important, and deserved much more research than they were getting.
When asked about the future effect of the rapid pace of innovation in silicon technology, Mead was emphatic that the 50 years of innovation provided us with a platform that is bringing about a new paradigm for the way we live and communicate. He likened its importance to the invention of print, and then of broadcast media. The new platform, which he believes is all about peer-to-peer connectivity, is just beginning to be explored. He cited eBay as an early example of the power of peer-to-peer. He was quick to point out that it was not just the exponential growth in silicon complexity that powers this new platform, but similar exponential growth in magnetic recording density, optical communications, and radio-based communications. Davidow added that there had been an exponential growth in knowledge, as well, which helped accelerate innovation across many disciplines.
As to the future of Moore's law, Mead was fairly sanguine. He made the point that throughout the 50 years, it has often appeared that technology innovation was about to "go off the edge of the world," but that some new technology or approach had always emerged to provide a path forward. He shared his simple vision of what is necessary to drive innovation -- that any time people know something is possible, and believe it is worth doing, they will make it happen. Perhaps that has been part of the secret of Moore's Law. Once it gained momentum as a predictor of progress, engineers had a target they believed they could achieve to aim for. Mead doesn't see anything changing that dynamic going forward. While he wasn't sure what specific technologies would be key in maintaining the rapid rate of progress, he is sure they will continue to emerge.
Mead summed up the energy and enthusiasm of Silicon Valley when he said that, over the first 50 years of Moore's Law, we have built an amazing platform of computation and networking capability, and we have only scratched the surface of how we can take advantage of it. He views that as our mission for the next 50 years.
[Image credits: Chemical Heritage Foundation(Opens in a new window), David Cardinal(Opens in a new window)]Tagged In
Integrated Circuit Intel Carver Mead Gordon Moore Computer History MuseumMore from Extreme
