Energy-Efficient Hardware Systems

The exponential growth in performance and storage capacity has been the key enabler for information technology for decades. However, the end of voltage scaling in semiconductor chips has made all computer systems, from mobile phones to massive data centers, energy limited. Moreover, new nanosystems enabled by emerging nanotechnologies provide unique opportunities for revolutionizing energy-efficient architectures through new transistor and memory technologies and their massive and fine-grained three-dimensional integration. These shifts motivate new system architectures and vertical co-design of hardware, system software, and applications. We look at new ways to design, architect, verify, and manage highly energy-efficient systems for emerging applications ranging from imaging and computer vision, machine learning, internet-of-things and big data analytics.

Examples include:

  • Hardware design for specialized accelerators and programming models for heterogeneous computing;
  • Scalable hardware verification and system validation;
  • Scalable architectures with thousands of computing elements and massive memory capacity;
  • Hardware architectures and systems software for cloud computing;
  • Architectures for nanosystems enabled by emerging technologies;
  • Robust and trustworthy architectures.

Faculty in Energy-Efficient Hardware Systems research area