@Scale 2018 Keynote: Glow: A community-driven approach to AI
In this keynote address, Jason Taylor, Vice President of Infrastructure at Facebook, announces the next steps in Facebook’s efforts to build a hardware ecosystem for machine learning (ML) through partner support of the Glow compiler. Cadence, Esperanto, Intel, Marvell, and Qualcomm Technologies Inc., a subsidiary of Qualcomm Incorporated, have committed to supporting Glow in future silicon products and technologies.
The rise of AI and machine learning has created a new set of bottlenecks for the network as workloads increase exponentially in size and scale. The traditional models for performance scaling are no longer enough to deal with these new sets of challenges. With an open source framework designed to be community driven, Glow allows partners to more rapidly design and optimize new silicon products for AI and ML processing by leveraging highly optimized compiler software.