CMOS semiconductor technology will face a dilemma in the 7nm process era in 2024, and graphene is expected to stand out as the best choice to replace this technology-this is based on the recent IEEE Custom Integrated Circuit Conference (CICC) held in California, USA The opinions expressed in a keynote speech.

　　"Graphene has shown many possibilities that will eventually replace silicon microchips, but I personally think that it will be another ten years at the earliest, and we will not see graphene applications appear until the silicon crystal material reaches its limit." Said James D. Meindl, professor of electronic and computer engineering at Georgia Institute of Technology. He is also the founding director of the University’s Nanotechnology Research Center, which has been dedicated to graphene research for five years.

　　In 2024, the silicon MOSFET will reach a bottleneck in terms of the length of the channel that can be manufactured and the thickness of the insulating gate that can be supported. Meindl quoted the International Semiconductor Technology Blueprint (ITRS) as a forecast.

　　Before becoming an alternative material to replace CMOS, graphene also faces many challenges. "We have to make billions of transistors on graphene sheets, but the number of transistors we can currently make is extremely small," Meindl said.

　　Researchers at the University of Manchester in the United Kingdom discovered this new material in a 2004 study. This Nobel Prize-winning research contribution is to find a new way to make a single layer of carbon atoms. Meindl recalled, "At the time, no one thought that this could be done, but it was actually just a starting point for applications that can be achieved by perfectly arranging a single layer of carbon atoms in a hexagonal honeycomb lattice."

　　So far, researchers have discovered at least two technologies that can be used to make graphene. They also made some "quite rough" effective transistors in this material.

　　Compared with copper interconnect technology, graphene transistors also have better electrical and thermal conductivity, and higher current carrying capacity. Graphene is also very attractive when used to make MEMS, Meindl said.

　　"So far, the most impressive graphene transistor has always been the RF transistor," such as amplifiers used for 500GHz analog signal applications, Meindl added. "The graphene switch is limited due to its leakage current and many other reasons. Make it more difficult to manufacture."

　　Meidl's laboratory is working on how to make 15nm line-width graphite ribbons, which can be used as a basis for building graphene switches as fast and effective as silicon switches. But the main challenge is to create a graphite ribbon with no damage to the edges, so as to avoid the degradation of the positive characteristics of the material.

　　Up to now, nearly 700 researchers from various engineering departments of Georgia Institute of Technology have visited Meindl's laboratory and explored graphene materials. "The interesting thing about our technology is that it is as broad as engineering, and almost as broad as physical science," he said.

　　The CICC conference agenda also includes a series of papers on various research topics from wired, wireless and optical communications to frequency, PLL, ADC and power components. It also includes discussions on special topics such as 3D chip stacks and biomedical technology.