Amazon Elastic Inference

Now you can reduce machine learning inference costs by up to 75% by using Amazon Elastic Inference (Amazon EI). This new accelerated compute service for Amazon SageMaker and Amazon EC2 enables you to add hardware acceleration to your machine learning inference in fractional sizes of a full GPU instance, so you can avoid over-provisioning GPU compute capacity. In this video, you'll also learn about the service's benefits and key features and see a brief demonstration.