AWS Lambda Now Runs GPU Workloads Without the Headache
AWS Lambda just dropped GPU support on Lambda Managed Instances using EC2 G-series hardware, letting you spin up AI inference, media processing, and scientific computing without wrestling with drivers or scaling. Available across all AWS Commercial Regions, you can go from code to GPU compute in minutes with pre-baked frameworks like PyTorch and vLLM.
source: [aws/whats-new]