How to do it...

1. Make sure the region you want to work in gives access to P2 or G3 instances. These instances include NVIDIA K80 GPUs and NVIDIA Tesla M60 GPUs, respectively. The K80 GPU is faster and has more GPU memory than the M60 GPU: 12 GB versus 8 GB. 

1. There are two options when launching an AWS instance. Option 1: You build everything from scratch. Option 2: You use a preconfigured Amazon Machine Image (AMI) from the AWS marketplace. If you choose option 2, you will have to pay additional costs. For an example, see this AMI at https://aws.amazon.com/marketplace/pp/B06VSPXKDX.
2. Amazon provides a detailed and up-to-date overview of steps to launch the deep learning AMI at https://aws.amazon.com/blogs/ai/get-started-with-deep-learning-using-the-aws-deep-learning-ami/.
3. If you want to build the server from scratch, launch a P2 or G3 instance and follow the steps under the Installing CUDA and cuDNN and Installing Anaconda and Libraries recipes.
4. Always make sure you stop the running instances when you're done to prevent unnecessary costs.