Get started with GPU accelerated instances - Amazon Elastic Compute Cloud

Get started with GPU accelerated instances

Fifth generation GPU accelerated instance types, such as those shown in the following list deliver the highest performance capabilities for deep learning and high performance computing (HPC) applications. Select the instance type link to learn more about its capabilities.

For a complete list of instance type specifications for accelerated instance types, see Accelerated computing in the Amazon EC2 Instance Types reference.

Software configuration

The easiest way to get started with fifth generation GPU accelerated instance types is to launch an instance from an AWS Deep Learning AMI that's preconfigured with all of the required software. For the latest AWS Deep Learning AMIs for use with GPU accelerated instance types, see the AWS Deep Learning Base GPU AMI (Ubuntu 20.04).

If you need to build a custom AMI to launch instances that host deep learning or HPC applications, we recommend that you install the following minimum software versions on top of your base image:

Software Instance type Minimum version
NVIDIA driver P5 530
P5e 550
CUDA P5, P5e 12.1
NVIDIA GDRCopy P5, P5e 2.3
EFA installer P5, P5e 1.24.1
NCCL P5, P5e 2.18.3
aws-ofi-nccl plugin P5, P5e 1.7.2-aws

We also recommend that you configure the instance to not use deeper C-states. For more information, see High performance and low latency by limiting deeper C-states in the Amazon Linux 2 User Guide. The latest AWS Deep Learning Base GPU AMI is preconfigured to not use deeper C-states.

For networking and Elastic Fabric Adapter (EFA) configuration see Maximize network bandwidth on accelerated computing instances with EFA.

Ubuntu 20.04 specific recommendations

The following recommendations for Ubuntu 20.04 help prevent unpredictable interface naming on boot:

  • Ensure you are running systemd 245.4-4ubuntu3.19 or later with the following command:

    systemd --version

  • Ensure you have configured GRUB:

    • Open the /etc/default/grub configuration file in a text editor.

    • Edit the GRUB_CMDLINE_LINUX_DEFAULT entry to include net.naming-scheme=v247.

    • Reboot your instance by running sudo update-grub.