Engineering machines that learn faster, compute more efficiently, operate safer, program easier, and integrate more responsibly (yes we do all that!)
A Few Recent Projects
Security from the Gates Up
Security and correctness issues continue to plague our systems, but increasingly it is the hardware at fault. Given that hardware is impossible to "patch", we need to new techniques that help us get our systems right the first time. The security techniques developed at UCSB are now deployed around the world and we continue to make important new discoveries weekly.
Hardware from Python
Building hardware can be as fun as coding again! PyRTL provides a collection of classes for register-transfer level design, simulation, tracing, and testing suitable for teaching and research. Simplicity, usability, clarity, and extensibility rather than performance or optimization is the overarching goal. We have used to for everything from OpenTPU to Zarfcore.
Brains have much to teach us about performing energy efficient computations. Our work on "race logic" has led to a class of neuro-inspired delay coded designs with orders of magnitude lower energy per computation as compared to prior approaches. There are more open questions than answers at this point but delay codes seem to unlock efficiency in exciting new ways.
A Couple of Recent Papers
- Weilong Cui, Yongshan Ding, Deeksha Dangwal, Adam Holmes, Joseph McMahan, Ali JavadiAbhari, Georgios Tzimpragos, Frederic T. Chong, and Timothy Sherwood. Charm: A Language for Closed-form High-level Architecture Modeling Proceedings of the International Symposium of Computer Architecture. (ISCA) June 2018. Los Angeles, Minnesota
- John Clow, Georgios Tzimpragos, Deeksha Dangwal, Sammy Guo, Joseph McMahan and Timothy Sherwood. A Pythonic Approach for Rapid Hardware Prototyping and Instrumentation Proceedings of the International Conference on Field-Programmable Logic and Applications (FPL) September 2017. Ghent, Belgium
- Joseph McMahan, Michael Christensen, Lawton Nichols, Jared Roesch, Sung-Yee Guo, Ben Hardekopf, and Timothy Sherwood. An Architecture Supporting Formal and Compositional Binary Analysis Proceedings of the 22th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), March 2017. Xian, China
- Advait Madhavan, Timothy Sherwood, and Dmitri Strukov. Race Logic: A Hardware Acceleration for Dynamic Programming Algorithms Proceedings of the International Symposium of Computer Architecture. (ISCA) June 2014. Minneapolis, Minnesota
"Advanced computer system design is not an island, but rather it sits between algorithms, machine learning, operating systems, compilers, circuits, networks, and security; by looking at computer architecture and embedded systems through the eyes of it's application we can enact the biggest change. This is the basic philosophy of our lab!"