Equinix reaches $100B valuation with Nvidia inference partnership

Equinix has reached a $100 billion market capitalization after announcing a partnership with Nvidia that positions the data center operator to serve the growing AI inference market, according to details first reported by CNBC.

The company's stock has climbed 33% in 2026, outpacing major technology stocks, as Equinix leverages its network of 281 colocation facilities across 77 metropolitan areas to address a different segment of AI infrastructure than hyperscalers building gigawatt-scale training centers.

Why it matters

As AI applications evolve beyond simple chatbots into complex agentic systems, inference workloads—where models make decisions on new data—are becoming more critical than training. McKinsey projects inference will represent half of all AI compute and 30% to 40% of total data center demand by 2030. Equinix's urban locations and distributed architecture address latency-sensitive inference needs that massive remote training facilities cannot, creating a defensible position in a market where hyperscalers are expected to spend over $5 trillion by 2030, according to Goldman Sachs Research.

The Nvidia deal structure

The partnership announced September 2, 2026, creates the Equinix Inference Exchange, which will launch in the first quarter of 2027. The program allows customers to run AI models on Together AI, an open-source cloud platform offering access to 200 models. Together AI will serve as the seller of record, billing end customers directly. Financial terms were not disclosed.

Nvidia CEO Jensen Huang joined Equinix CEO Adaire Fox-Martin at a San Francisco event, noting that Equinix's distributed facility locations allow companies to be "close to where the action is, where all the sensors are."

Maryam Zand, an Equinix vice president running AI ecosystem strategy, described the offering as an "inference platform as a service" that enables customers to connect across multiple clouds and providers while optimizing costs. Equinix facilities are configured for Nvidia's B300 Blackwell Ultra GPUs, with some liquid-cooled locations supporting the newer Vera Rubin chips.

A different infrastructure strategy

Unlike neoclouds such as CoreWeave that build massive facilities for AI training, Equinix has maintained its focus on smaller data centers—most under 100 megawatts—positioned near urban centers as network interconnection hubs. This approach has drawn criticism from short seller Jim Chanos, who told CNBC in May that legacy data center companies are "very capital-intensive" and "don't grow that fast."

Yet the numbers suggest Equinix's model generates consistent profits. In its latest quarter, Equinix reported revenue of $2.63 billion, up 16% year-over-year, with net income of $477 million. CoreWeave, by comparison, posted $2.58 billion in revenue—more than double the prior year—but lost $626 million in the same period.

Vlad Galabov, host of the AIDC Debate podcast and longtime data center analyst, said Equinix is "super diversified" with general purpose compute across a wide client base serving over 10,500 customers. "They are not exposed to an AI bubble risk," Galabov said. "That means you're missing out on some of the boom. It's just inevitable. High risk, high return."

Equinix also announced Equinix Fabric One, a connectivity service designed to simplify networks operating across multiple clouds and AI models.

CNBC reporters Isabel O'Brien and Katie Tarasov first reported these developments.