TIDEZINE.

NVIDIA PAIR: Turning That Idle PC at Home into a Free AI Compute Pool

NVIDIA unveiled PAIR, a free open-source tool, at IFA 2026. It distributes AI compute workloads to idle computers on your local network, so your main machine doesn't have to shoulder the entire GPU load alone.

NVIDIA PAIR: Turning That Idle PC at Home into a Free AI Compute Pool

One computer is busy gaming while another sits idle in the living room—NVIDIA sees this everyday scene as a compute problem. Personal AI Router (PAIR), unveiled by the company at IFA 2026, is a free, open-source tool with a simple mission: find idle computers on your local network and offload AI compute tasks to them.

NVIDIA notes in its official announcement that more than half of U.S. households own two or more computers, most of which sit idle for the majority of the time. That's the premise behind PAIR—and the waste it's designed to fix.

When AI agents tackle complex tasks, they typically break the work into multiple subtasks running in parallel—but those subtasks often still end up competing for the same GPU, slowing everything down. PAIR's job is to spread those subtasks across other idle computers in the house, rather than piling them all onto one machine to queue up.

NVIDIA's example: sorting a cluttered inbox. A user asks an AI agent to prioritize urgent emails; the agent splits the job into multiple sub-agents, and PAIR routes those sub-agents to idle PCs on the network. Meanwhile, the user's own machine keeps running smoothly for gaming or work, unburdened by the AI workload.

PAIR is currently in beta, supporting Windows, macOS, and Linux, with both a graphical interface and a command-line option. Hardware requirements call for an NVIDIA GeForce RTX 20-series or newer GPU, an NVIDIA RTX PRO workstation GPU (Turing architecture or later), or an NVIDIA DGX Spark; on the Apple side, an M4 chip or newer is required.

NVIDIA has not yet announced an official launch date or a complete feature list for PAIR. All public details currently trace back to the company's official announcement at IFA 2026.

Follow what interests you

Start with a brand tag or a section to build your own reading list.

Related

Adobe Overhauls Photoshop and Premiere's Core Architecture, Doubling Speed with NVIDIA RTX Spark Optimization
Tech

Adobe Overhauls Photoshop and Premiere's Core Architecture, Doubling Speed with NVIDIA RTX Spark Optimization

Adobe says Photoshop, Premiere, and Substance 3D have been re-architected to run AI tasks and editing workflows up to twice as fast on computers equipped with NVIDIA RTX Spark.

iPhone Duo Pre-orders Pushed to November, First Batch Still Shipping October 23
Tech

iPhone Duo Pre-orders Pushed to November, First Batch Still Shipping October 23

Apple's new iPhone Duo sold out within minutes of pre-orders opening. New orders are now reportedly delayed to a November 16-23 shipping window, while the first batch of store and customer orders remains on track for October 23.

Kojima Productions Teams Up with CASETiFY on Death Stranding Accessories Designed by Yoji Shinkawa
Tech

Kojima Productions Teams Up with CASETiFY on Death Stranding Accessories Designed by Yoji Shinkawa

Kojima Productions and CASETiFY have launched a collaboration celebrating their 10th and 15th anniversaries, featuring phone cases designed by Yoji Shinkawa and Death Stranding character accessories, released on October 15.

ChatGPT Can Now Read Your Chrome Tabs to Compare Prices, Fill Forms, and Draft Emails for You
Tech

ChatGPT Can Now Read Your Chrome Tabs to Compare Prices, Fill Forms, and Draft Emails for You

OpenAI has updated the ChatGPT Chrome extension so it can read multiple open tabs to compare and act on their content—from comparing sofa specs to filling out forms—and it's available to all paid plans.

Tesla Unveils Next-Gen Roadster: Upside-Down Ceiling Stunt, Jump-Ramp Launch, and a $250K Entry Fee
Tech

Tesla Unveils Next-Gen Roadster: Upside-Down Ceiling Stunt, Jump-Ramp Launch, and a $250K Entry Fee

Tesla debuted its next-generation Roadster at SpaceX's Texas test site, suspending the car upside-down with a chassis suction system before pulling off a jump-ramp stunt powered by cold-gas thrusters. Production is slated for 2028.

One Model, Two Computers: Meta Open-Sources Muse Spark Open, and the Entry Fee Is 1TB of Memory
Tech

One Model, Two Computers: Meta Open-Sources Muse Spark Open, and the Entry Fee Is 1TB of Memory

Meta has released Muse Spark Open, an 800-billion-parameter open-source model optimized for Apple Silicon. But to actually run it locally, you'll need a Mac Studio setup that could cost up to $38,000.