News
Key Takeaways
- Meta launches Meta Compute to scale AI data centers.
- The initiative is led by Santosh Janardhan, Daniel Gross and Powell McCormick.
- Senior IT and energy stakeholders face rising power needs and supply shifts.
Meta is betting big on AI infrastructure, unveiling Meta Compute to build tens of gigawatts of data center capacity this decade. According to CEO Mark Zuckerberg, the January 12, 2026 announcement marks a top-level initiative targeting hundreds of gigawatts over time.
Santosh Janardhan, who oversees technical architecture and the global datacenter fleet, will lead the effort alongside Daniel Gross, who will manage long-term capacity strategy and supplier partnerships. Newly appointed President and Vice Chairman Dina Powell McCormick will lead government and sovereign partnerships for infrastructure financing.
Table of Contents
- Meta Compute: Infrastructure Scope and Commitments
- Meta’s AI Buildout Accelerates Across Chips, Cloud and Agents
- The Physical Bottlenecks Slowing AI’s Expansion
- Meta at a Glance
Meta Compute: Infrastructure Scope and Commitments
Meta committed up to $72 billion in 2025 capital spending and signed 20-year power purchase agreements with three Vistra nuclear plants in the US heartland. The tech company is also developing projects with two firms building small modular reactors.
With Meta Commute, the company plans to build tens of gigawatts this decade, and hundreds of gigawatts or more over time. The team also aims to partner with governments and sovereigns to build, deploy, invest in and finance Meta's infrastucture.
"I'm looking forward to working closely with Daniel, Santosh, Dina and their teams to scale Meta Compute and deliver personal superintelligence to billions of people around the world," said Zuckerberg.
Related Article: xAI vs Mistral vs Meta: Who Has the Strongest AI Moat?
Meta’s AI Buildout Accelerates Across Chips, Cloud and Agents
Meta has aggressively expanded its AI capabilities through major acquisitions and infrastructure investments throughout 2025 and early 2026. The company acquired a 49% non-voting stake in Scale AI for $14.8 billion in June 2025, hiring CEO Alexandr Wang to lead its new Meta Superintelligence Labs.
In September, Meta announced plans to acquire chip startup Rivos to strengthen in-house semiconductor development, and in December bought AI startup Manus for over $2 billion to integrate autonomous agent technology.
On infrastructure, Meta signed a $14.2 billion agreement with CoreWeave through 2031 and secured a six-year, $10 billion cloud computing deal with Google, while agreeing to the OneGov initiative with the US General Services Administration.
The company released Llama 4 in April 2025 in three tiers — Scout (17B), Maverick (34B) and a still-delayed 200B+ Behemoth — offering 256,000-token context and multimodal capabilities. Meta has also doubled down on AI wearables as sales of smartglasses tripled this year.
The Physical Bottlenecks Slowing AI’s Expansion
AI's explosive growth has slammed into hard physical limits — energy, water and land constraints now threaten enterprise viability.
AI supercomputing performance has doubled roughly every nine months, but power consumption for large-scale systems has doubled every 12 months, outpacing gains in energy efficiency. The largest AI supercomputers now consume energy on the scale of small cities. xAI's Colossus, for example, draws up to 300 megawatts, enough to power over 250,000 US homes.
Webinar
On demand
Rebrand. Migrate. Optimize. How to Do It All (Without Slowing Down)
Cresta leveled up site speed, design flexibility and marketer sanity (in record time). Find out how.
Watch Now
Webinar
On demand
From Manual to Magical: How AI Transforms CX Teams
Learn how to replace manual support processes with automation that actually delivers.
Watch Now
Webinar
On demand
How to Build a Solid Knowledge Foundation for AI Success
See how leading brands keep their AI honest, compliant and actually helpful.
Watch Now
Webinar
On demand
Fix the Content Bottleneck: Build a Better WebOps Strategy
Content stalled? Dev overloaded? You’re not the only one. Learn how streamlined WebOps bridges the publishing gap.
Watch Now
Webinar
On demand
Beyond Storage: Smarter Content, Bigger Impact with DAM + AI
Discover how the DAM + AI duo makes content smarter, stronger and more accessible.
Watch Now
Webinar
On demand
Agentic AI Playbook: Real-World Customer Service Use Cases You Can Deploy Now
Boost self-service by 30% and slash call volume by 63% with agentic AI.
Watch Now
Data centers collectively consumed nearly a quarter of Virginia's entire electricity supply in 2024. The US Department of Energy projects data centers could require 12% of total US electricity.
Facing increasing power requirements, major operators including Amazon, Google and Microsoft have announced partnerships with nuclear power companies. These infrastructure demands are driving interest in large language models that can deliver more capability with greater efficiency.
Related Article: AI's Voracious Appetite for Land, Water and Power Is Your Next Big Business Risk
Meta at a Glance
Meta, founded in 2004 as TheFacebook, Inc., offers social media platforms including Facebook, Instagram, WhatsApp, Messenger and Threads, alongside hardware like Meta Quest VR headsets and Ray-Ban Meta AI glasses. The tech giant is currently headquartered in Menlo Park, California.
