Reservoir’s Algorithms Team focuses on the applied mathematics and software that can achieve asymptotic speedups in our high performance computing projects.  This provide both differentiating capabilities in our products and services, and also helps us shape our compiler and systems work toward accelerating algorithms that are on the frontier of HPC.

A substantial amount of Reservoir’s algorithms work has been directed toward problems in sensing and signal processing.  Reservoir’s algorithms enable both qualitative new capabilities and the ability to implement these algorithms with substantially lower size weight and power (SWAP).  Examples of our work are Generalized Compressive Processing (GCP) and the Sparse Multidimensional FFT (sMFFT).

Other parts of our algorithms work have been directed toward analytics and machine learning, to provide the ability to solve problems in these areas (for discovery, inference, and learning) with asymptotic performance and scale increases.

For additional examples of some of Reservoir’s algorithms work, please see the relevant publications.