- Amazon Web Services (AWS) has launched Amazon Elastic Compute Cloud (EC2) Capacity Blocks for ML, enabling customers to reserve access to NVIDIA H100 Tensor Core GPUs for AI-related tasks.
- The product provides access to clusters ranging from one to 64 instances, each with eight GPUs, for defined periods.
- Customers can book resources for up to 14 days in one-day increments and up to eight weeks in advance.
- The dynamic pricing of the offering depends on supply and demand. The service provides cost certainty for customers and helps AWS optimise resource use.