Bottlerocket now supports NVIDIA Multi-Instance GPU (MIG) for Kubernetes workloads
News
AWS has announced Bottlerocket's support for NVIDIA Multi-Instance GPU (MIG) in Kubernetes environments, enabling more efficient GPU resource utilization.
- Allows partitioning NVIDIA GPUs into multiple GPU instances on Kubernetes nodes
- Enables running multiple workloads simultaneously on a single GPU
- Provides hardware-level memory and fault isolation between workloads
- Optimizes GPU resource allocation for tasks like machine learning inference
- Available in all commercial and AWS GovCloud (US) Regions with compatible GPU instances
This feature helps system administrators maximize GPU performance and efficiency by enabling more granular GPU resource sharing.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Mar 5
2025
2025
Bottlerocket now supports AWS Neuron accelerated instance types
Oct 23
2024
2024
NVIDIA GPU Time-slicing Now Available for Bottlerocket to Enhance AI/ML Workload Efficiency
Dec 16
2024
2024
Bottlerocket now supports Elastic Fabric Adapter for AI/ML and HPC workloads
Apr 29
2025
2025
Amazon EKS Hybrid Nodes now supports Bottlerocket
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.