Cloud & Compute Infrastructure
AWS Lambda added G-series GPU instances for Event Functions on Lambda Managed Instances, supporting inference, media processing, and rendering.
· ComputeLabs Research · from the November 18, 2026 edition
AWS announced support for G-series GPU instances for Event Functions running on Lambda Managed Instances. The scope is this specific Lambda execution offering, rather than an announcement that every Lambda function configuration now includes GPU access.
AWS identifies machine-learning inference, media processing, and rendering as supported workload categories. The announcement concerns GPU-backed cloud instances, not the sale of physical GPUs or a disclosed expansion measured in data-center power capacity.
Lambda manages the underlying instance lifecycle, GPU drivers, and autoscaling. AWS also describes sharing GPU utilization across concurrent requests, although the supplied excerpt provides no benchmark, concurrency limit, or specific G-series instance-model list.
- AWS Lambda
- G-series GPU instances
- Event Functions
- Lambda Managed Instances

