Paper Title: Evolving Storage for a New Generation of AI/ML

Paper Abstract: AI/ML is not new, but innovations in ML model development have made it possible to process data at unprecedented speeds. Data scientists have used standard POSIX file systems for years. However, as the scale and need for performance has grown, many face new storage challenges. Samsung has been working with customers on new ways of approaching storage issues with object storage designed for use with AI/ML. This talk will present how software and hardware are evolving to allow unprecedented performance and scale of storage for Machine Learning.

Paper Author: Somnath Roy, Principal Engineer, System architecture, Samsung Semiconductor

Author Bio: My name is Somnath Roy and presently I am working as Principle engineer in Samsung Memory Solution Lab. I have more than 18 years of experience in different areas of ever evolving storage stack. I also have several years of experience on building and performance architecting private cloud solutions built upon distributed Object Storage. My accomplishments includes multiple patents in storage fields and contribution to some of the popular distributed object store like CEPH, MinIO. Presently, at Samsung we are working on building a distributed disaggregated object storage solution with industry leading performance that can potentially serve very high READ bandwidth requirement of AI training workload at multi PB scale environment.