Basic Research Needs for Microelectronics: Report of the Office of Science Workshop on Basic Research Needs for Microelectronics, October 23 – 25, 2018
Cherry A. Murray, Supratik Guha, Dan Reed, Gil Herrera, Kerstin Kleese van Dam, Sayeef Salahuddin, James S. Ang, Thomas M. Conte, Debdeep Jena, Robert Kaplar, Harry A. Atwater, Rick Stevens, Dushan Boroyevich, William J. Chappell, Tsu‐Jae King Liu, Justin Rattner, M. S. Witherell, Khurram K. Afridi, Simon S. Ang, Jon Bock, Srabanti Chowdhury, Suman Datta, K. R. Evans, Jack Flicker, M.a. Hollis, N. M. Johnson, K. A. Jones, Peter M. Kogge, Sriram Krishnamoorthy, Matthew Marinella, Todd Monson, Sreekant Narumanchi, Paul R. Ohodnicki, R. Ramesh, Michael L. Schuette, John Shalf, Shadi Shahedipour‐Sandvik, J. A. Simmons, Valerie Taylor, Tom Theis, E. Colby, Robinson E. Pino, Andy Schwartz, Katie Runkles, Joseph E. Harmon, Michele Nelson, Vicki Skonicki2018 · DOI
For over five decades, the progress of computing was defined by Moore’s Law—a predictable doubling of transistor density every two years. This era of "free lunch" scaling enabled a drop in computing costs by eight orders of magnitude, revolutionizing every facet of modern society. However, we have reached a critical juncture. Physical limits at the nanoscale and the economic realities of chip fabrication have brought traditional scaling to a halt. To continue advancing, we must abandon the modular, linear development cycles of the past in favor of a holistic "co-design" approach that integrates materials science, device physics, and system architecture.
The Challenge: Beyond the Von Neumann Bottleneck
The primary obstacle to next-generation computing is energy efficiency. In current architectures, the energy required to move data between memory and processors—the "von Neumann bottleneck"—now significantly exceeds the energy used for the computation itself. As we enter the age of Artificial Intelligence (AI) and massive scientific data sets, this inefficiency has become unsustainable.
The Department of Energy (DOE) faces immediate pressures from its own scientific missions. Future facilities like the High-Luminosity Large Hadron Collider (HL-LHC) are expected to generate raw data at rates of 1 petabyte per second (PB/s). Processing and storing this data within reasonable power budgets requires a 10,000x improvement in system-level energy efficiency. Similarly, the transition to a resilient, clean electricity grid requires a revolution in power electronics to manage bidirectional energy flows from intermittent renewable sources.
A New Paradigm: Holistic Co-design
The report proposes a Holistic Co-design Framework as the solution. Traditionally, materials scientists developed new materials, passed them to device engineers, who then handed components to architects. Co-design replaces this linear path with a multi-directional feedback loop.
Conceptual illustration of the co-design framework, bridging the gap between atomic-scale material properties and large-scale device architectures.
In this framework, application requirements (top-down) dictate the necessary properties of new materials, while new physical phenomena (bottom-up), such as spintronics or negative capacitance, inspire entirely new non-von Neumann architectures. This requires the development of advanced Electronic Design Automation (EDA) tools capable of simulating systems from the atomic level up to the software stack.
Priority Research Directions
The workshop identified several key areas where fundamental research can break current bottlenecks:
Overcoming the "Boltzmann Tyranny": Standard silicon transistors are limited by a minimum voltage swing required to switch. "Beyond-CMOS" devices utilizing tunneling or ferroelectric materials are needed to operate at much lower voltages.
Compute-in-Memory: By moving logic directly into the storage elements, we can eliminate the energy-intensive data movement that plagues current CPUs.
Reimagining Interconnects: The "last centimeter" of wiring on a chip is a major bottleneck. Research into using photons, plasmons, or bio-inspired "artificial axons" could provide the necessary bandwidth density (1 Pb/s/mm) at a fraction of the energy cost.
Ultra-Wide Bandgap (UWBG) Semiconductors: Materials like diamond and aluminum nitride are essential for the next generation of power electronics, potentially shrinking massive power converters to the size of a suitcase and enabling a more agile, cyber-secure electricity grid.
Quantitative Goals and Future Outlook
The report sets ambitious benchmarks for the research community:
Energy Efficiency: A target of 100 attojoules (aJ) per single computation, representing a 10,000x improvement over current high-performance computing systems.
Thermal Management: Moving beyond air cooling to integrated microfluidics and nanoscale thermal engineering to handle the extreme heat densities of 3D-stacked chips.
Material Discovery: "Industrializing" the discovery process through high-throughput computational screening to move past the current "artisanal" pace of development.
The transition to this new era of microelectronics is not merely a technical challenge but a sociological one. It requires bridging the cultural silos between materials scientists and software engineers. Without a sustained, coordinated investment in these basic research needs, the U.S. risks losing its leadership in scientific discovery and global economic competitiveness.
Summary updated
Abstract
Moore’s Law — realized via a combination of device physics advances, technology investments, and economic returns— allowed the number of transistors on a chip to double roughly every two years for over five decades. During those five decades, the cost of a unit of computing dropped by eight orders of magnitude. Those declines and the associated computing advances have dramatically affected every aspect of society, from science and technology through business and health to national security. Today, further computing advances are limited by the economics of chip fabrication and the physical limits on transistors, electrical interconnects, and memory elements at the nanoscale. If we are to evolve new generations of computing systems over the next decades that are both faster and more energy efficient, a complete reorganization of the science and technology underlying computing is needed. Within the Department of Energy’s (DOE’s) mission, which includes a pivotal and historical role in the evolution of computing, the needs are critical as well. Advanced computing and simulation underpin all aspects of DOE missions in energy, the environment, and national security, requiring energy-efficient computing beyond the exascale range. Edge computing, low-power computing technologies, and computers optimized for artificial intelligence are key to next-generation scientific facilities for high-energy physics research, as well as neutron and x-ray facilities. Finally, microelectronics will also play a major role in reshaping the U.S. electricity grid: from its current state to one that is cleaner, more efficient, cyber-secure, and resilient to widespread events, both natural and manmade. Realizing this vision will require advancements in both system architecture and power electronics. To enable continued advances in computing technologies, a fundamental rethinking is needed of the science behind the materials, synthesis and placement technologies, architectures, and algorithms. This cannot be modular and linear, as it has been in the past. Rather, these advances must be developed collectively, in a spirit of co-design, where each scientific discipline informs and engages the other to achieve orders of magnitude improvements in system-level performance. To explore these challenges, the Office of Science convened a Basic Research Needs Workshop for Microelectronics in October 2018 and charged workshop participants to conduct a thorough assessment of the scientific issues associated with advanced microelectronics technologies for applications relevant to the DOE mission. The workshop examined research relevant to the extension of complementary metal oxide semiconductor (CMOS) and beyond CMOS technologies; however, topics of direct relevance to quantum information science and quantum computing were outside the scope of this workshop.
BibTeX
@misc{murray2018basic,
author = {Cherry A. Murray and Supratik Guha and Dan Reed and Gil Herrera and Kerstin Kleese van Dam and Sayeef Salahuddin and James S. Ang and Thomas M. Conte and Debdeep Jena and Robert Kaplar and Harry A. Atwater and Rick Stevens and Dushan Boroyevich and William J. Chappell and Tsu‐Jae King Liu and Justin Rattner and M. S. Witherell and Khurram K. Afridi and Simon S. Ang and Jon Bock and Srabanti Chowdhury and Suman Datta and K. R. Evans and Jack Flicker and M.A. Hollis and N. M. Johnson and K. A. Jones and Peter M. Kogge and Sriram Krishnamoorthy and Matthew Marinella and Todd Monson and Sreekant Narumanchi and Paul R. Ohodnicki and R. Ramesh and Michael L. Schuette and John Shalf and Shadi Shahedipour‐Sandvik and J. A. Simmons and Valerie Taylor and Tom Theis and E. Colby and Robinson E. Pino and Andy Schwartz and Katie Runkles and Joseph E. Harmon and Michele Nelson and Vicki Skonicki},
title = {{Basic Research Needs for Microelectronics: Report of the Office of Science Workshop on Basic Research Needs for Microelectronics, October 23 – 25, 2018}},
year = {2018},
doi = {10.2172/1616249},
}