From 8d82b60c9566de9572d2880f38848ac93bb82f0f Mon Sep 17 00:00:00 2001 From: Greg Leeds Date: Tue, 11 Oct 2016 15:46:50 -0500 Subject: [PATCH] Updating EC2 GPU section to reflect new instance types --- README.md | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 7c2887f..1e8e947 100644 --- a/README.md +++ b/README.md @@ -699,7 +699,12 @@ EC2 - Aside from bootstrapping, you should manage keys yourself on the instances, assigning individual keys to individual users or services as appropriate. - Avoid reusing the original boot keys except by administrators when creating new instances. - How to avoid sharing keys; how to add individual ssh keys for individual users. -- **GPU support:** You can rent GPU-enabled instances on EC2. There are [two instance types](http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using_cluster_computing.html). Both sport an NVIDIA card (K520, 1536 CUDA cores and M2050, 448 CUDA cores). +- **GPU support:** You can rent GPU-enabled instances on EC2 for use in machine learning or graphics rendering workloads. + - There are [three generations](http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/using_cluster_computing.html) of GPU-enabled instances available: + - Third generation P2 series offers NVIDIA K80 GPUs in 1, 8 and 16 GPU configurations targeting machine learning and scientific workloads. + - Second generation G2 series offers NVIDIA K520 GPUs in 1 or 4 GPU configurations targeting graphics and video encoding. + - First generation CG1 instances are still available in some regions in a single configuration with a NVIDIA M2050 GPU. + - 🔹Typical GPU workloads are often a good fit for [Spot Instances](#ec2-cost-management). ### EC2 Gotchas and Limitations