Amazon EC2 Capacity Blocks for ML can now be shared across multiple AWS accounts via AWS Resource Access Manager. This capability allows organizations to pool reserved GPU resources, optimizing utilization across different teams and workloads without additional charges.
By facilitating coordinated use of reserved capacity, this feature lowers total infrastructure expense and improves operational efficiency for machine learning deployments. It is generally available in supported AWS Regions and intended to help organizations manage ML capacity investments more flexibly.
Source & further reading
Amazon EC2 capacity blocks for ML can be shared across multiple accounts ↗Source published 2026-02-05. This report was prepared with AI assistance from the linked primary-source material. Claims and limitations should be checked against the original announcement or research.