Roadmap to the Semiconductor Future

By Rebooting Computing Portal Staff

The worldwide semiconductor industry ( is one of the major drivers of economic development and technological progress, with sales of order $300 Billion per year. This has become far too big for any one company or country to dominate, In order to continue to improve performance into the future, planning for the next decade and beyond is essential. In the past, the industry could depend on Moore's Law ('s_law), with doubling of device speed and density every one to two years. While the end of Moore's Law has been predicted for the past 20 years, factors such as power density finally seem to be preventing further progress.

The International Technology Roadmap for Semiconductors (ITRS) represents the major chip manufacturers across the world, and prepares a planning report (the "Roadmap") every two years (with updates between the reports) for future directions in integrated circuit manufacturing for computers and consumer electronics. The 10-page Executive Summary of the 2013 Edition is available at This includes an assessment of the near term (through 2020), with some projections of the longer term (through 2028).

While traditional trends of reducing the scale of CMOS transistors are continuing, they are slowing down dramatically, requiring 3D architectures to further increase device densities. Longer term projections are looking toward non-CMOS memory technologies, such as magnetic RAM (MRAM), phase-change memory (PCM), Resistive RAM (ReRAM), and ferroelectric RAM (FeRAM). These must be integrated with CMOS manufacturing and systems. Furthermore, the report surveys emerging research devices and materials such as carbon nanotubes and graphene.

So while the traditional Moore's Law scaling is ending, the semiconductor industry plans to continue to offer improved cost-effective and energy-efficient system performance for diverse industries including computers, telecommunications, consumer electronics, and automotive.

The ITRS Roadmap, with its emphasis on large-scale manufacturing in the semiconductor industry, is somewhat complementary to the more open approach of the Rebooting Computing Working Group, but the impressive track record established by earlier versions of the Roadmap show that it provides an important guidepost to any assessment of future computing.