NVIDIA releases PAIR software to pool multiple PCs for local AI

NVIDIA has released a free, open-source beta software tool called Personal AI Router (PAIR) designed to pool the computing power of multiple computers on a local network to run AI models. The software automatically detects compatible devices and distributes AI agent workloads across them, addressing performance bottlenecks caused by single-GPU competition for tasks.

In company benchmarks, PAIR reduced the time for a representative five-agent task from 18 minutes to 8 minutes and 48 seconds by spreading the workload across three devices. The system is designed for privacy, as all data processing occurs within the local network rather than in external clouds.

PAIR is compatible with Windows, macOS, and Linux. Supported hardware includes NVIDIA GeForce RTX 20 Series GPUs and newer, RTX PRO workstation GPUs, DGX Spark systems, and Apple M4 silicon. The tool works alongside platforms such as Ollama and LM Studio and can adapt dynamically as devices join or leave the network.