The Gadget Guy reviews the latest technology for the government and explores related trends and hot topics.
Back in 1965, Intel co-founder Gordon E. Moore observed the density of transistors on an integrated circuit doubles approximately every two years. He observed this as having happened over the prior seven years or so, and predicted it should continue for at least another decade.
Nearly 50 years later, computing technology has more or less kept up with this pattern fairly regularly. But delays in the development of the latest transistor creation method have threatened to bring Moore’s Law to a halt.
Microchips are made with a process called “photolithography.” A light-sensitive layer coats the entire silicon chip, then is burned away with light shown through a pattern to leave the transistors. Obviously, the limiting factor here is the intricacy of the pattern used, as well as the wavelength of the light (the shorter, the better). Right now, photolithography uses light in the ultraviolet range, but to go much smaller, manufacturers will have to switch to the extreme ultraviolet range, with wavelengths less than a tenth of what they are using now.
The problem that has reared its head since the idea of using EUV was developed lies in producing light in that range that is bright enough to do the job. The dimmer the light used, the longer each wafer has to be exposed, and this would hamper production considerably. To produce chips at a rate that would allow the manufacturers to stay in business, about 250 watts of EUV light would be needed. They are working toward this goal, if a little slowly.
At the SPIE Advanced Lithography conference this past February, light-source maker Cymer announced it had gotten light levels up to 55W. Skip Miller, the director of strategic marketing of EUV machine manufacturer ASML, says many of the major kinks in EUV light production have been worked out and the company is more confident the light source brightness can be boosted all the way to 250 W. “We can now build systematically a plan to get to 250 W,” Miller says.
There are many technical obstacles to getting production-level EUV microchip construction going. But given recent developments, many experts are confident we will have them churning out by 2015. This would allow Moore’s prediction to remain more or less correct.