Limits in Sight for Moore's Law?

Back in the 1960s, electronics industry pioneer Gordon Moore of Intel Corp. made a prediction, now immortalized as "Moore's Law", to the effect that transistor density on integrated circuits would continue to double every two years. This prediction of exponential development has largely proven valid, as ever-shrinking transistor size has resulted in increased performance and decreased cost for chips. Partly as a result of this "Law", chips have become the basic component of computers, mobile phones, and just about every other modern electronic device. Recently, Intel's staff have been revisiting the subject and Moore, now Intel's Chairman Emeritus of the Board, spoke of extending Moore's Law at the International Solid-State Circuits Conference (ISSCC), earlier this year. Four researchers co-authored a technical paper published in Proceedings of the IEEE entitled "Limits to Binary Logic Switch Scaling--A Gedanken Model". The research paper posited that an upper limit will be reached by chipmakers attempting to shrink the size of transistors and some estimates predicts this limit will be reached in about 18 or 19 years. The basic issue is that chip technology is approaching atomic dimensions in terms of its components. A transistor essentially consists of two physically separated regions between which electrons may pass through a controlled gate. Effectively allowing or preventing passage of the electrons through the gate is fundamental to the functioning of the transistor. As the distance of the physical separation drops below five nanometers, a phenomenon called electron tunneling will begin to occur: electrons will simply pass through the separation on their own, regardless of any attempt at control, because the two regions will be extremely close. With the loss of effective gate control, the present transistor design will simply not work. Some approaches to the postulated ceiling focus on new materials, including strained silicon, silicon-on-insulator, and double- and triple-gate transistors. Other methods being explored are the more efficient use of electrons or simply making bigger chips. For a report of Moore's message, see: http://intel.com/labs/features/eml02031.htm. For a report on the research paper, see: http://news.com.com/2100-7337-5112061.html. Summary prepared by: Peter Wang