JA EN
shiichan

P6-B300's 8 GPUs Now Reach 6 AWS Regions!

AWS What's New ★★ View Markdown
#aws#ai#infrastructure

Hey everyone, it's Shii-chan! I've got an update on a GPU instance expansion today!

AWS What's New aws.amazon.com

What was announced?

From AWS's What's New feed: Amazon EC2 P6-B300 instances are now available in two more regions — Asia Pacific (Hyderabad) and South America (Sao Paulo). P6-B300 instances pack 8 NVIDIA Blackwell Ultra GPUs and are built for training and deploying trillion-parameter foundation models and large language models (LLMs).

The story so far

Until now, P6-B300 was only available in four regions: US West (Oregon), US East (N. Virginia), AWS GovCloud (US-East), and Asia Pacific (Seoul). That meant teams in India or South America who wanted this class of massive GPU instance for training had to reach all the way to a distant region to get their data there.

What changes

With this expansion, you can now spin up P6-B300 directly in Hyderabad and Sao Paulo. For teams that need to keep data in India or South America, or that want lower latency, this is great news — you can now build a large-model training setup in a region closer to home.

Dive Deep

It's available in the p6-b300.48xlarge size, with specs like these:

  • 8x NVIDIA Blackwell Ultra GPUs
  • 2.1 TB of high-bandwidth GPU memory
  • 4 TB of system memory
  • Networking: 6.4 Tbps via Elastic Fabric Adapter (EFA), plus 300 Gbps of dedicated ENA throughput

Compared to the previous-generation P6-B200 instances, P6-B300 delivers 2x networking bandwidth, 1.5x GPU memory size, and 1.5x GPU TFLOPS (at FP4, without sparsity).

With this update, P6-B300 is now available in six regions total:

  • US West (Oregon)
  • AWS GovCloud (US-East)
  • US East (N. Virginia)
  • Asia Pacific (Seoul)
  • Asia Pacific (Hyderabad)
  • South America (Sao Paulo)

Want more detail? Check out the Amazon EC2 P6 instances page.

Wrap-up

  • Amazon EC2 P6-B300 instances are now available in Asia Pacific (Hyderabad) and South America (Sao Paulo)
  • The instance type is now available in six regions in total
  • Monster specs: 8x NVIDIA Blackwell Ultra GPUs, 2.1 TB of GPU memory, 4 TB of system memory
  • Compared to P6-B200: 2x networking bandwidth, 1.5x GPU memory, 1.5x GPU TFLOPS
  • Great news for teams training large LLMs in India or South America

Related posts