Software Frameworks for NVMe-based Computational Storage
Dr. Stephen Bates, Eideticom
Computational Storage promises to increase the efficiency of computer systems and address the concerns of CPU scaling. It aims to do this by offloading key compute tasks to more efficient offload engines and to reduce data movement by moving processing closer to storage.
However, in order to achieve this promise new software frameworks are required. In this talk we look at several examples of software frameworks for an NVM Express-based Computational Storage Device (CSx). We discuss the advantages and disadvantages of these approaches as they relate to performance, scalability and ease of integration into existing applications.
Finally, we show how these Software Frameworks can be applied to several existing applications including an HPC parallel filesystem and the HDFS framework in Hadoop.
Stephen is an expert in performance storage, persistent and non-volatile memory, computer networking, signal processing and error correction coding. He is one of the drivers of the NVMe-based Computational Storage efforts within SNIA and NVM Express. He holds a PhD from The University of Edinburgh and is a Senior Member of the IEEE.