June 2025
Intermediate to advanced
338 pages
9h 17m
English
This chapter will cover strategies to maximize graphics processing unit (GPU) (https://aws.amazon.com/what-is/gpu/) efficiency in K8s when deploying GenAI applications as GPU instances are very expensive and often underutilized. We will also cover GPU resource management, scheduling best practices, and partitioning options such as Multi-Instance GPU (MIG) (https://www.nvidia.com/en-us/technologies/multi-instance-gpu/), Multi-Process Service (MPS) (https://docs.nvidia.com/deploy/mps/index.html), and GPU time-slicing (https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/gpu-sharing.html). Finally, we’ll discuss monitoring GPU performance, balancing workloads across ...
Read now
Unlock full access