Every two to three years, the number of transistors on a computer chip doubles, but the silicon that has driven this progress for decades is running out of room to shrink. This project tackles a fundamental bottleneck in the semiconductor industry. For over fifty years, making transistors smaller has boosted computing power, but silicon’s physical limits now prevent further miniaturisation without sacrificing performance. Without a replacement, the steady improvements in speed, efficiency, and cost that underpin everything from mobile phones to factory control systems would stall. The University of Glasgow team is developing transistors made from compound semiconductors—materials that can switch faster and use less power than silicon. Working with Freescale Semiconductors, they aim to deliver a prototype technology ready for large-scale manufacture by around 2016. The research spans atomic-level crystal growth, computer simulations to optimise design, and fabrication techniques compatible with existing silicon factories. If successful, this work could extend the trajectory of Moore’s Law into the next decade, keeping integrated circuits—and the devices, infrastructure, and medical equipment that rely on them—on their current path of improvement. The technology may also prove useful beyond digital logic, in sensors, imaging, and communications.
View original technical description
The semiconductor industry is one of the largest on the planet, with a turnover of more than $200 billion each year. Integrated circuits produced by the semiconductor industry are found inside all modern electronic appliances and products including mobile phones, cars, medical diagnostic equipment, controlling the safe operation of factories and public transportation systems and powering the Internet . In short, they are vital to modern life in the 21st Century. Since the invention of the transistor in 1957, manufacturers such as AMD, Intel, IBM and Freescale have been successful in developing ever more complex integrated circuits by making the individual transistors smaller and finding ways to combine more of them together on a single chip. The result has been a regular increase in the computational and processing capability of integrated circuits by doubling the number of transistors in each circuit every 2-3 years. Currently the most advanced integrated circuits contain hundreds of millions of transistors, each of which is 1/10,000 of the diameter of a human hair in size. Until now, this increase in capability has resulted from making smaller silicon-based transistors, however fundamental limits imposed by the properties of silicon are now being reached so that alternative materials need to be considered. In the view of all the major manufacturers mentioned above, a strong candidate to enable continued performance improvements for the industry are compound semiconductors. The Nanoelectronics Research Centre at the University of Glasgow is one of the world leaders in compound semiconductor transistor technology. For the last 3 years we have been working closely with Freescale Semiconductors to develop such transistor technologies which will, by around 2016, be ready for large scale manufacture as required for continued integrated circuit performance improvement at that time. This 3.8M, 3 year project, is focussed on delivering prototype compound semiconductor transistor technology, capable of being scaled up to large volume manufacture, with the required performance to deliver the types of processing and control functions required by integrated circuits in 2016. In addition to applications using digital logic such as microprocessors, this technology is expected to be of more general use, in areas such as sensors and photonics for medial, safety, imaging and communications applications. Five teams from the University of Glasgow will participate in the research which will cover compound semiconductor growth techniques controlled to atomic level precision; electrical, chemical and structural characterisation of the semiconductor materials the transistors fabricated from them, again on atomic lengthscales; powerful computer simulation to optimise transistor design; developing compound semiconductor processing techniques compatible with existing, well understood silicon-based methods; and building prototype transistors to show that the performance requirements are being met. Together, this project will deliver key information and understanding which will enable the semiconductor industry to continue to be one of the most successful on the planet in the coming decades.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know