Workshops on Extreme Scale Design Automation (ESDA) Challenges and Opportunities for 2025 and Beyond

R. Iris Bahar, Alex K. Jones, Srinivas Katkoori, Patrick H. Madden, Diana Marculescu, Igor L. MarkovarXiv 2020 · arXiv

Cover of the CCC white paper on Extreme Scale Design Automation.

Executive Summary

This report summarizes the strategic findings of a Computing Community Consortium (CCC) workshop series focused on the future of Electronic Design Automation (EDA). As the industry moves into the post-Dennard scaling era, the EDA community must navigate extreme-scale silicon complexity, integrate emerging post-CMOS technologies, and expand its methodologies into new domains like synthetic biology and cyber-physical systems.

The Crisis in Design Productivity

The EDA industry has long been the engine behind Moore’s Law, providing the tools necessary to manage billions of transistors. However, the field has reached a critical juncture. The end of classical "Dennard scaling"—where smaller transistors yielded automatic improvements in power and performance—has created a widening "productivity gap."

Currently, semiconductor fabrication capabilities (such as sub-10nm nodes and 3D Integrated Circuits) are outstripping the ability of design tools to handle them efficiently. This technical challenge is compounded by commercial stagnation and a conservative R&D environment within the industry, making it difficult to attract new academic talent to the field.

A Strategic Shift in Methodology

To address these challenges, the report advocates for a fundamental shift in how EDA research is approached:

  • Big Data and Machine Learning: Moving away from traditional algorithmic heuristics toward data-centric frameworks. These frameworks can better manage the massive datasets generated during modern system-level design and validation.
  • New Abstractions: Creating higher-level mathematical models to shield designers from the physical complexities of emerging technologies, such as quantum computing or carbon-based electronics.
  • System-Level Integration: Fostering tighter collaboration between computer architecture, software engineering, and hardware design to optimize entire systems rather than individual components.

Key Strategic Insights

The "Design Automation of Things" (DAoT)

EDA’s primary success has been creating abstractions that allow designers to build complex chips without needing to be physicists. This paradigm can be exported to other fields. The "Design Automation of Things" (DAoT) suggests applying EDA principles to synthetic biology, wearable medical devices, and smart grids.

Optimizing Legacy Nodes

While the industry focuses on sub-10nm nodes, the majority of IoT and embedded devices rely on mature "legacy" nodes (e.g., >80nm). There is significant research potential in squeezing extreme power efficiency and cost savings out of these nodes rather than viewing them as solved problems.

Modeling Post-CMOS Technologies

Emerging technologies like memristors and spintronics often face cycles of hype. To evaluate these pragmatically, the EDA community must develop accurate device models and simulation tools now. This allows for system-level comparisons before significant investments are made in fabrication.

Actionable Recommendations

The workshops reached a consensus on several key actions for funding agencies and the research community:

  1. Invest in Extreme-Scale EDA: Focus on big-data research for design, verification, and reliability at the most advanced nodes.
  2. Support Post-CMOS Abstractions: Fund programs to develop the metrics and benchmarks needed for automated design using non-CMOS technologies.
  3. Launch DAoT Initiatives: Use targeted funding to apply design automation to Cyber-Physical Systems (CPS) and hardware security.
  4. Workforce Revitalization: Use MOOCs and summer schools to draw computer science talent back to the EDA field.

Challenges and Future Outlook

The path forward is hindered by the conservative nature of the EDA industry, which is dominated by a few large companies that often prefer acquiring startups over funding high-risk internal R&D. Furthermore, the risk of disrupting time-sensitive design flows often prevents the adoption of superior academic algorithms.

The report calls for "patient capital"—sustained, long-term funding for high-risk exploration—and the establishment of new community-building events to unify researchers under the DAoT umbrella.

The Computing Community Consortium (CCC) logo.

Summary updated

Abstract

Integrated circuits and electronic systems, as well as design technologies, are evolving at a great rate -- both quantitatively and qualitatively. Major developments include new interconnects and switching devices with atomic-scale uncertainty, the depth and scale of on-chip integration, electronic system-level integration, the increasing significance of software, as well as more effective means of design entry, compilation, algorithmic optimization, numerical simulation, pre- and post-silicon design validation, and chip test. Application targets and key markets are also shifting substantially from desktop CPUs to mobile platforms to an Internet-of-Things infrastructure. In light of these changes in electronic design contexts and given EDA's significant dependence on such context, the EDA community must adapt to these changes and focus on the opportunities for research and commercial success. The CCC workshop series on Extreme-Scale Design Automation, organized with the support of ACM SIGDA, studied challenges faced by the EDA community as well as new and exciting opportunities currently available. This document represents a summary of the findings from these meetings.
BibTeX
@misc{bahar2020workshops,
  author = {R. Iris Bahar and Alex K. Jones and Srinivas Katkoori and Patrick H. Madden and Diana Marculescu and Igor L. Markov},
  title = {{Workshops on Extreme Scale Design Automation (ESDA) Challenges and Opportunities for 2025 and Beyond}},
  year = {2020},
  eprint = {2005.01588},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2005.01588},
}
Figure