Application-specific near-memory data transformations to improve both storage efficiency and parallel execution performance.
A genome sequencing accelerator with new hardware-friendly automaton theory.
Data approximation on optical links for high bandwidth and low latency communication on HPCs.