Maximizing GPU utilization with Preemptive Scheduling and Transparent Checkpoint

Speaker: Dr. Charles Fan, co-founder and CEO of MemVerge.

Talk Abstract:
As generative AI continues to redefine what's possible, GPUs have become the backbone of its computational infrastructure. Yet despite their power and cost, GPU utilization across teams and projects often remains frustratingly low. For AI professionals tasked with deploying and managing generative AI applications, this inefficiency translates into higher costs, slower iteration cycles, and limited scalability.

A core issue lies in the challenge of sharing GPU resources across multiple workloads with differing priorities. Traditional job schedulers lack the flexibility and intelligence to adapt to the dynamic nature of real-world AI deployment environments.

In this talk, I’ll introduce a next-generation GPU scheduling paradigm built for the realities of AI operations. This approach allows you to assign priorities to different projects and users, enabling high-priority generative workloads to preempt lower-priority tasks. Combined with transparent checkpointing, preempted jobs can be paused and seamlessly resumed—dramatically improving GPU utilization without sacrificing progress or reliability.

This talk is ideal for ML engineers, DevOps teams, and infrastructure leads looking to get more out of their GPU investments while scaling generative AI apps more efficiently.

Speaker Bio:
Charles Fan is co-founder and CEO of MemVerge. Prior to MemVerge, Charles was the CTO of Cheetah Mobile leading its global technology teams, and an SVP/GM at VMware, founding the storage business unit that developed the Virtual SAN product. Charles also worked at EMC and was the founder of the EMC China R&D Center. Charles joined EMC via the acquisition of Rainfinity, where he was a co-founder and CTO.
Charles received his Ph.D. and M.S. in Electrical Engineering from the California Institute of Technology, and his B.E. in Electrical Engineering from the Cooper Union.

Register online at:
https://us02web.zoom.us/meeting/register/yrGbMu2eTw2PZe2KS3VDeg

7:00-7:05 pm Welcome
7:05-7:45 pm Presentation
7:45-8:00 pm Q&A