AI news
Microsoft and Amazon Launch "Decentralized AI Grid" (DAIG) for Edge Inference
Date: June 29, 2026
Microsoft and Amazon have announced a strategic joint venture to establish the Decentralized AI Grid (DAIG), a global infrastructure project designed to shift AI inference from massive data centers to the network edge. This initiative utilizes a novel "Swarm Inference" protocol, allowing complex model queries to be processed across a distributed fabric of enterprise and consumer hardware.
| Feature | Specification / Impact |
|---|---|
| Primary Objective | Offload 40% of centralized inference traffic to the edge |
| Performance | 60% reduction in round--trip latency for real--time agents |
| Sustainability | Estimated 30% reduction in data center cooling energy requirements |
| Security | Hardware--level TEE (Trusted Execution Environment) encryption |
Industry analysts view this as a direct response to the energy constraints of centralized superclusters. By leveraging existing global hardware, DAIG aims to democratize access to low--latency AI while bypassing the current power grid bottlenecks facing traditional data centers. "The future of AI is not just in the cloud, but in the mesh of devices surrounding the user," stated the joint press release.