Pricing and usage model updates for HAQM EC2 instances accelerated by NVIDIA GPUs
AWS announces two updates to HAQM EC2 instances accelerated by NVIDIA GPUs.
Availability of savings plans for HAQM EC2 P6-B200 instances which were available at launch only through EC2 Capacity Blocks for ML.
Reduced pricing for HAQM EC2 P5 and P5en instances and HAQM EC2 P4d and P4de instances: The following pricing reductions apply to On Demand pricing beginning June 1, 2025 and to Savings Plan purchases effective after June 4, 2025: P5 - up to 45% reduction, P5en - up to 26% reduction, and P4d and P4de - up to 33% reduction. The percentages apply to instances running HAQM Linux, slightly smaller reductions apply to instances running other operating systems. To provide increased accessibility to reduced pricing, we are also making at-scale, On-Demand capacity available for:
- P4d in Asia Pacific (Seoul), Asia Pacific (Sydney), Canada (Central), Europe (London) Regions
- P4de in US East (N. Virginia) Region
- P5 in Asia Pacific (Mumbai), Asia Pacific (Tokyo), Asia Pacific (Jakarta), South America (São Paulo) Regions
- P5en in Asia Pacific (Mumbai), Asia Pacific (Tokyo), Asia Pacific (Jakarta) Regions
The new pricing reflects AWS's commitment to making advanced GPU computing more accessible while passing cost savings directly to customers. To learn more about the updated pricing please consult the EC2 pricing page.