Skip to main content
Trending
Corporate earningsAi infrastructureEnergy marketsElection litigationAviation safetyCrypto securityAi governanceEnterprise aiPlatform liabilityAvian influenzaClinical trialsFederal debtPresidential runoffRpg developmentSpace debrisAir defenseAutonomous vehiclesBudget crisisBug bountiesEnergy m&aForeign investmentFranchise restructuringGene therapyPandemic treatyTranslation pathwaysUn peacekeepingWeapons testingWeight-space programmingAi copyright lawArchaeological toxicology

full timeline

Microsoft to Unveil Maia 300 AI Chip, Negotiates TSMC Capacity for 1M Units to Cut Nvidia Reliance

Microsoft has introduced its next-generation Maia 300 silicon, designed specifically to handle AI inference workloads across its global data centers. The custom hardware will roll out in 2027 as part of a multi-gigawatt infrastructure expansion aimed at reducing reliance on external suppliers.

Factlen Newsroom· Updated

Follow Microsoft to Unveil Maia 300 AI Chip, Negotiates TSMC Capacity for 1M Units to Cut Nvidia Reliance in the Factlen app and get an alert when there is news. Get the app

Microsoft is preparing its global data centers for a multi-gigawatt deployment of custom artificial intelligence silicon.From: Microsoft Unveils Maia 300 AI Chip to Power Cloud Inference Across Multi-Gigawatt Fleet

Share this story

Charts, maps and images

A single gigawatt of electrical capacity is roughly equivalent to the power required by 750,000 homes.From: Microsoft Unveils Maia 300 AI Chip to Power Cloud Inference Across Multi-Gigawatt Fleet ·
The computational burden in data centers is rapidly shifting from model training to daily inference tasks.From: Microsoft Unveils Maia 300 AI Chip to Power Cloud Inference Across Multi-Gigawatt Fleet ·
Illustration: Securing high-yield fabrication capacity is a critical prerequisite for a hardware deployment of this magnitude.From: Microsoft Unveils Maia 300 AI Chip to Power Cloud Inference Across Multi-Gigawatt Fleet ·
The company is negotiating with TSMC to secure advanced manufacturing capacity for 2027.From: Microsoft to Unveil Maia 300 AI Chip, Negotiates TSMC Capacity for 1M Units to Cut Nvidia Reliance ·
Lowering the cost of inference could provide massive relief for AI labs running large-scale models.From: Microsoft to Unveil Maia 300 AI Chip, Negotiates TSMC Capacity for 1M Units to Cut Nvidia Reliance ·

Timeline

October 2026

  1. Microsoft Unveils Maia 300 AI Chip to Power Cloud Inference Across Multi-Gigawatt Fleet

August 2026

  1. Microsoft to Unveil Maia 300 AI Chip, Negotiates TSMC Capacity for 1M Units to Cut Nvidia Reliance