Standardizing Hyperscaler Requirements for GPU Management and Acceleration
Open Compute Project via YouTube
Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Learn about GPU standardization efforts in hyperscale data centers through this collaborative panel discussion featuring experts from AMD, Google, Meta, Microsoft, and NVIDIA. Explore how industry standards like MCTP, SPDM, PLDM, and Redfish are being implemented to streamline GPU device management in hyperscale environments. Discover the progress made in developing standards-based profiles for GPU management, designed to accelerate the implementation and adoption of new GPU designs. Gain insights into the challenges posed by the current lack of standardization in GPU management and varying hyperscaler requirements. Examine the improvements achieved in GPU management over the past year through the Open Compute Project's initiatives.
Syllabus
Standardizing Hyperscaler Requirements for Accelerators
Taught by
Open Compute Project