NVIDIA Open Sources PAIR to Unite Macs and RTX PCs Into a Single AI Cluster
NVIDIA has launched the Personal AI Router (PAIR) software in an open-source beta, functioning as a local network scheduler that directs AI requests to available computers.

Automatic Device Discovery and Request Assignment: Direct Benefits for Ollama and LM Studio
On compatible hardware, including M4+ Macs, NVIDIA RTX PCs, and DGX Spark, PAIR provides a unified interface for local AI tools like Ollama and LM Studio. This allows users to distribute multiple independent AI tasks across idle devices within a home or office network.
Designed as a local request scheduler, PAIR automatically detects connected computers and assigns tasks to available resources. Since models continue to run via Ollama or LM Studio on the target device, users avoid configuring each machine individually. If a device is busy, PAIR redirects requests to other paired nodes. For instance, AI agents processing multiple documents can assign independent tasks to different computers. NVIDIA clarifies that this distributes concurrent workloads and does not accelerate individual AI responses by combining computing power.
Performance Test: Task Splitting Cuts Time from 18 Minutes to 8 Minutes 48 Seconds
NVIDIA’s demo showed that using Ollama with the Qwen 3.6 35B A3B model in Hermes, splitting tasks among five AI sub-agents reduced the average completion time for a single RTX Spark laptop from 18 minutes to 8 minutes and 48 seconds after adding a DGX Spark and an RTX 5090.
Related article
UK AI Startup Emulate Launches Fundraising With Target of Up to $700 Million and Valuation of $3.7 Billion
On Wednesday, September 16, UK-based AI startup Emulate announced a new funding round targeting up to $700 million, achieving a post-money valuation of approximately $3.7 billion. The round was led by Index Ventures and Lightspeed Venture Partners. E
AI founders weigh open vs. closed models at TechCrunch Disrupt 2026
For AI startups, selecting a model is no longer just a one-time architectural choice.Open models are getting better. Frontier APIs keep improving. Models can be tailored for specific tasks, and some companies are building products that use multiple m
Moonshot AI’s Kimi K2.5, the first trillion-parameter multimodal model, retires at month’s end
Moonshot announced via Kimi’s official Weibo that its inaugural trillion-parameter multimodal model, Kimi K2.5, will be retired by the end of this month. Launched and open-sourced in January, K2.5 stands as Kimi’s most versatile model to date, featur
Related Special Topic Recommendations
Comments (0)
0/500
NVIDIA has launched the Personal AI Router (PAIR) software in an open-source beta, functioning as a local network scheduler that directs AI requests to available computers.

Automatic Device Discovery and Request Assignment: Direct Benefits for Ollama and LM Studio
On compatible hardware, including M4+ Macs, NVIDIA RTX PCs, and DGX Spark, PAIR provides a unified interface for local AI tools like Ollama and LM Studio. This allows users to distribute multiple independent AI tasks across idle devices within a home or office network.
Designed as a local request scheduler, PAIR automatically detects connected computers and assigns tasks to available resources. Since models continue to run via Ollama or LM Studio on the target device, users avoid configuring each machine individually. If a device is busy, PAIR redirects requests to other paired nodes. For instance, AI agents processing multiple documents can assign independent tasks to different computers. NVIDIA clarifies that this distributes concurrent workloads and does not accelerate individual AI responses by combining computing power.
Performance Test: Task Splitting Cuts Time from 18 Minutes to 8 Minutes 48 Seconds
NVIDIA’s demo showed that using Ollama with the Qwen 3.6 35B A3B model in Hermes, splitting tasks among five AI sub-agents reduced the average completion time for a single RTX Spark laptop from 18 minutes to 8 minutes and 48 seconds after adding a DGX Spark and an RTX 5090.
UK AI Startup Emulate Launches Fundraising With Target of Up to $700 Million and Valuation of $3.7 Billion
On Wednesday, September 16, UK-based AI startup Emulate announced a new funding round targeting up to $700 million, achieving a post-money valuation of approximately $3.7 billion. The round was led by Index Ventures and Lightspeed Venture Partners. E
AI founders weigh open vs. closed models at TechCrunch Disrupt 2026
For AI startups, selecting a model is no longer just a one-time architectural choice.Open models are getting better. Frontier APIs keep improving. Models can be tailored for specific tasks, and some companies are building products that use multiple m
Moonshot AI’s Kimi K2.5, the first trillion-parameter multimodal model, retires at month’s end
Moonshot announced via Kimi’s official Weibo that its inaugural trillion-parameter multimodal model, Kimi K2.5, will be retired by the end of this month. Launched and open-sourced in January, K2.5 stands as Kimi’s most versatile model to date, featur





Home






