Data and AI releases

2,002 this month · 180 companies

Amazon ElastiCache now supports distributed KV caching for self-hosted LLM inference

Generally availableNot yet verified

Amazon ElastiCache for Valkey launched fully managed distributed KV caching for self-hosted LLM inference engines like vLLM to lower context computation costs and response latency.

LLMsPerformance

Read on aws.amazon.com

Also shipped Nov 18, 2026AWS in November 2026 →

More from Amazon Aurora and DynamoDBEvery Amazon Aurora and DynamoDB release →