Nvidia announced its Personal AI Router (PAIR), a free software tool designed to connect idle computers – desktops, laptops, and even Apple M4 chips or newer – to perform local AI inference tasks. The tool aims to utilize underused computing power within homes for tasks like running AI agents.
PAIR is not a hardware device but open-source software that discovers and connects compatible PCs on a network. It’s designed to work with Nvidia GeForce GPUs (RTX 20-series and newer, RTX Pro, and DGX Spark systems) as well as Apple’s M4 chips and newer. Nvidia product manager Seth Schneider explained that the system operates by leveraging computers when they are not actively being used, preventing interference with other tasks. This distributed system can then process requests in parallel, reducing bottlenecks and adapting as devices join or leave the network, even if a user begins gaming on one of the connected PCs.
According to Schneider, a household with a DGX Spark desktop, an RTX Spark laptop, an RTX 5090 laptop, a gaming desktop, and a MacBook Pro could have approximately 165 teraflops of underutilized compute power. He stated, “It’s truly a treasure trove of free tokens just sitting in homes today,” even factoring in electricity costs. Nvidia secures PAIR by requiring a six-digit pairing code and utilizing mTLS (Mutual Transport Layer Security) to create an encrypted communication line between devices.
The Nvidia PAIR beta is currently available for Windows, Linux, and macOS. Nvidia also announced that Perplexity Portable Computer, Hermes Agent, and OpenClaw will offer simplified local setup with Nvidia GPUs on Windows, allowing users to get AI agents running with fewer manual configurations. Nvidia envisions a typical user having one laptop and one gaming PC to utilize the system.
Read the original coverage
💬 Comments
📜 Comment Policy