AI infra push spans cloud, industry, and U.S. hubs
The global AI infrastructure race is broadening across multiple fronts — from neocloud partnerships and enterprise AI to rapid data center buildouts. In the U.S., CoreWeave has signed a landmark $14.2 billion agreement with Meta, supplying Nvidia’s latest GB300 systems in a deal that cements the neocloud’s role at the center of next-generation compute while diversifying its revenue base beyond Microsoft. In Europe, Paris-based Mistral AI is deepening its push into Industry 4.0, signing a company-wide AI deployment deal with Stellantis while securing a €1.7 billion Series C round led by ASML, with which it is co-developing industrial-grade LLMs trained on proprietary manufacturing data. And in Texas, Crusoe has powered on the first phase of its Abilene campus, an Oracle-backed project designed as a blueprint for future large-scale AI hubs, already running Nvidia GB200 racks within just over a year of breaking ground. Together, these moves highlight how AI infra is no longer just about hyperscale compute — it is now expanding into enterprise-specific applications, regional sovereignty, and new models of speed-to-market. Here’s how the next phase of AI infra is taking shape.
Juan Pedro Tomas
Editor
RCRTech
AI Infrastructure Top 3
CoreWeave’s $14B Meta deal: CoreWeave secures a $14.2 billion deal with Meta for Nvidia-powered compute, diversifying beyond Microsoft and underscoring surging demand for AI infrastructure as hyperscalers and neoclouds scale capacity through 2031.
Mistral AI at Stellantis: Paris-based Mistral AI has signed a major deal with automaker Stellantis to deploy customized industrial AI models, following a €1.7 billion funding round led by ASML, with which it is also co-developing industrial LLMs.
Abilene AI campus goes live: Crusoe has energized its Abilene AI data center campus with Oracle, bringing Nvidia GB200 racks online within a year of groundbreaking and setting a new benchmark for U.S. AI infrastructure speed and scale.
AI-Powered Telecom Infrastructure
Supermicro, in collaboration with NVIDIA, delivers AI-powered infrastructure tailored for telcos, enhancing operational efficiency, network management, and customer experiences. Explore now
AI Today: What You Need to Know
Meta strengthens AI chip plans: Meta is acquiring chip startup Rivos, bolstering its in-house semiconductor push as it seeks to reduce reliance on Nvidia and cut AI infrastructure costs amid soaring demand for advanced computing.
ASUS expands AI infra: ASUS leverages its Taiwania II supercomputing expertise to deliver sovereign GPU clusters and eco-friendly data centers in the Middle East, partnering with Nvidia, Weka, and Schneider to drive regional AI sovereignty and scale.
Semtech readies 1.6T optics: Semtech is sampling 200G/channel VCSEL and TIA chipsets, enabling power-efficient multimode 1.6T optical interconnects. Hyperscalers gain lower-cost, short-reach options as AI clusters expand toward 2026 volume manufacturing.
Microsoft shifts to in-house chips: CTO Kevin Scott says Microsoft plans to prioritize its Maia AI accelerators and Cobalt CPUs in data centers, reducing reliance on Nvidia and AMD while advancing full-stack AI system design.
123NET boosts Michigan GPU hub: 123NET expanded its Southfield DC1 site with 4MW high-density power, liquid and air cooling, and Detroit Internet Exchange peering, creating Michigan’s first integrated AI-ready colocation hub.
DOE opens Oak Ridge to AI: The U.S. DOE invited bids to build AI data centers at Oak Ridge, Tennessee. Proposals will require new power generation and storage, with decisions due December 1.
Taiwan rejects U.S. chip demand: Taipei pushed back against U.S. calls to shift 50% of semiconductor production to American soil, citing its “silicon shield” as vital to security while sustaining TSMC’s central global role.
Nscale adds $433M funding: UK-based Nscale closed a $433 million Pre-Series C SAFE days after its $1.1B Series B. Backers include Dell, Nvidia, and Nokia, reinforcing its sovereign AI infra expansion push.
Cerebras raises $1.1B: Cerebras secured $1.1 billion Series G at an $8.1 billion valuation, led by Fidelity. Funds will expand U.S. chip manufacturing, data center capacity, and next-gen AI supercomputers.