AWS Machine Learning Inference Optimization in SageMaker Neo improves latency and throughput for ML models deployed across various hardware architectures.

https://aws.amazon.com/sagemaker/neo/