Accelerate large-scale AI applications with the new Amazon EC2 P6-B300 instances
AWS News Blog
This article announces the general availability of Amazon EC2 P6-B300 instances, powered by NVIDIA Blackwell Ultra GPUs for large-scale AI applications.
- 2x more networking bandwidth and 1.5x more GPU memory than previous generation
- 8x NVIDIA B300 GPUs with 2.1TB total GPU memory per instance
- 6.4 Tbps Elastic Fabric Adapter networking for efficient GPU cluster communication
- Ideal for training trillion-parameter models and Mixture of Experts architectures
- Supports GPUDirect Storage for up to 1.2 Tbps throughput to Lustre file systems
- Available in US West (Oregon) via Capacity Blocks for ML and Savings Plans
P6-B300 instances provide balanced compute, memory, and networking for distributed AI training at scale.
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
May 15
2025
2025
New Amazon EC2 P6-B200 instances powered by NVIDIA Blackwell GPUs to accelerate AI innovations
Nov 18
2025
2025
Amazon EC2 P6-B300 instances with NVIDIA Blackwell Ultra GPUs are now available
Sep 2
2025
2025
Unlocking next-generation AI performance with Dynamic Resource Allocation on Amazon EKS and Amazon EC2 P6e-GB200
Jul 9
2025
2025
New Amazon EC2 P6e-GB200 UltraServers accelerated by NVIDIA Grace Blackwell GPUs for the highest AI performance
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.