Our lab is currently installing SUMMIT, our next generation supercomputer with ~27,000 NVIDIA Voltas, and we are preparing for the next system in 2021, which is expected to have a more complex architecture. So, our Future Technologies team is investigating LLVM programming solutions for emerging parallel/heterogeneous/soc architectures. We work at the cutting edge of computer architecture, codesigning and evaluating technologies for future systems including heterogeneous computing with GPUs and FPGAs, nonvolatile memory (e.g., 3D Xpoint memory), and even quantum and neuromorphic computing. Our work is funded by US Department of Energy, DARPA, and others.
We are growing and have several immediate opportunities for R&D in Clang/LLVM including a postdoc position, a software engineer position, and a research staff member position. These positions will focus on extensions to Clang/LLVM to support heterogeneity in compute and memory, parallelism in the IR, and specialization for domain-specific systems-on-a-chip. Our team’s products are typically open source, and we contribute to Clang/LLVM and publish at popular venues. Non-US citizens are welcome (and we are happy to help with visa applications).
If you are interested in applying or learning more, please contact me at email@example.com.