- 最新
- 投票最多
- 评论最多
Not sure how you can see x86 for g5g.xlarge since its Graviton based ARM processor. But its available in us-east-1 :
Hi,
Why don't you try to use this script to see if where the g5g instances are available for your account? It checks for all regions and AZs,
https://gist.github.com/trestletech/f93d32e04c601b0584c0ce1a421e9948
Best,
Didier
I believe Sandy was not able to find the Deep Learning AMI for G5g. I have provided links to the AMI below. I don't think they are listed under the official Amazon AMI releases.
Click the links on this page to see the ami-ids for the various Graviton Deep Learning AMIs: https://docs.aws.amazon.com/dlami/latest/devguide/overview-cpu.html
You will also find links for getting started with the Graviton GPU DLAMI.
On further investigation, it appears all the DLAMIs for Graviton have been deprecated. I can not find them anymore. You may need to build you own. Here's a blog post showing how to do it: https://mirzabilal.com/deep-learning-with-aws-graviton2-nvidia-tensor-t4g-for-as-low-as-free-with-cuda-12-2-56d8457a6f6d
I found an AMI from NVIDIA. Search for "NVIDIA GPU-Optimized AMI (ARM64)" under "AWS Marketplace AMIs". G5g should be supported and there doesn't appear to be any charge from NVIDIA. It does not include the ML framework, only: Ubuntu Server 20.04 NVIDIA Driver 515.48.07 Docker-ce 20.10.17 NVIDIA Container Toolkit 1.10.0-1 NVIDIA Container Runtime 3.10.0-1 AWS Command Line Interface (CLI) Miniconda 4.13.0 JupyterLab 3.4.3 and other Jupyter Core packages NGC CLI 3.0.0 Git, Python3-PIP More details here: https://docs.nvidia.com/ngc/ngc-deploy-public-cloud/ngc-aws/index.html
Also check this blog post, Using Graviton G5g instances with Nvidia NGC Containers: https://medium.com/@wayne_toh/using-graviton-g5g-instances-with-nvidia-ngc-containers-16fcd9bb3062
I believe Sandy was not able to find the Deep Learning AMI for G5g. I have provided links to the AMI below. I don't think they are listed under the official Amazon AMI releases.