Sampa (historically meant to stand for safe multi-processing architectures) is a collection of research projects aimed at making computer systems faster, more reliable, more energy efficient, supporting and motivating new applications. We are part of the Computer Science and Engineering Department at the University of Washington.
The nature of the research problems we are interested in is inter-disciplinary and so are we. Our projects include innovation in architecture, operating systems, compilers and programming languages.
Congratulations to Brandon Lucia on his CGO paper on data provenance in multithreaded programs (the last chapter of his thesis); to Thierry, Mark Wyse, Adrian and Jacob on their HPCA paper on FPGA-based neural networks; and to Mike Ringenburg and Adrian on their ASPLOS paper on debugging and monitoring quality in approximate programs! Great news!!
Check out the documentation and code for ACCEPT, our approximate compiler framework.
Brandon Lucia, one of our proud alumni, is moving to CMU as an assistant professor! Congratulations, Brandon!
Our group will have two papers appear in OOPSLA 2014! One on locality analysis for global-address-space parallel code and one on symbolic execution for multithreaded programs. Great stuff, congratulations, Tom B, and Brandon H!
Tom Bergan and Mike Ringenburg have recently finished their PhDs! Mike is now an Architect at Cray, and Tom will join Google Seattle in the Fall. Congratulations to both!
Our paper on analog neural networks for approximate computing was accepted to ISCA'14. Great collaboration with GATech, UT Austin and MSR!
Our work is supported by the National Science Foundation, the state of Washington, the C-FAR Center for Future Architectures, and generous gifts from Microsoft, Intel, Google, Qualcomm and NetApp.