Innovation & Tech

Kyndryl Lands Argentum AI in $630M, 20-Year Boulder Data Center Deal

Argentum AI signed a 20-year, $630 million deal for 10 MW at Kyndryl's Boulder site, deploying NVIDIA B300 GPUs for inference under a power-first model.

By Nathan Brooks

3 min read

Kyndryl’s Boulder Site Welcomes Argentum AI as Flagship Anchor Tenant
Kyndryl’s Boulder Site Welcomes Argentum AI as Flagship Anchor TenantAkela999 / pixabay

What's News

  • Argentum AI and Kyndryl signed a 20-year, $630 million data center collaboration on November 17, 2026
  • Initial capacity is 10 megawatts at Kyndryl's Boulder, Colorado facility, with NVIDIA B300 systems for AI inference
  • Deployment is already underway in a fully Tier III environment built for current-generation GPU power density
  • Argentum reports 77,000 GPUs contracted, 344 MW of contracted power and 13 global sites under exclusivity
  • Argentum becomes the Boulder facility's flagship anchor tenant

Kyndryl and Argentum AI signed a 20-year, $630 million data center collaboration on November 17, 2026, anchored at Kyndryl's Boulder, Colorado facility with an initial 10 megawatts of capacity. Argentum will deploy NVIDIA B300 systems at the site for AI inference workloads.

Deployment is already underway in a fully Tier III environment. The existing Boulder facility was built for the power density of current-generation GPU systems, which lines up with the hardware Argentum is installing. For practitioners, that means no construction lag — the companies say the enterprise-grade infrastructure is intended to run demanding AI applications at scale and give Argentum's customers a faster path to high-performance inference capacity.

Argentum becomes the facility's flagship anchor tenant, and Kyndryl gets exactly what a landlord wants: a long-dated commitment from a tenant that actually fills the building. "Our agreement gives us a unique opportunity to put Kyndryl capacity to work in an innovative way," said Jamie Rutledge, president of Kyndryl U.S. Rutledge said Kyndryl's mission-critical capabilities and decades of operational and services experience would support Argentum as it creates capacity for customers.

What does the power-first model change?

Argentum originates, finances and operates GPU clusters, delivering dedicated bare-metal GPU capacity to AI model builders, inference providers and enterprises. Its deployment sequence matters: secure site control and power first, buy GPUs second. Only then does it coordinate financing, OEM relationships, cluster design and operational partners. Argentum promotes this power-first sourcing as a faster alternative to construction-dependent buildouts.

The Kyndryl deal supplies the facility and power-capacity side of that model. Argentum handles GPU deployment and the customer-facing service on top. The approach is designed around customer demand, with capacity matched to workload requirements rather than speculative building.

Financing is the other half of the story. Argentum describes financing structures built for contracted AI deployment, intended to scale capacity without waiting for capital to be assembled deal by deal. The 20-year Kyndryl commitment fits that model — a defined infrastructure requirement paired with Argentum's broader financing and cluster-operation role.

How much of the stack can customers outsource?

Customers choose their own level of handoff, across three tiers:

  • Hardware-as-a-Service — physical infrastructure, power and connectivity, while the customer runs the OS, software and operations
  • Bare-Metal-as-a-Service — adds operating-system delivery, break-fix work and maintenance
  • Platform-as-a-Service — extends Argentum's responsibility through AI application management and reporting

How big is Argentum's wider pipeline?

The Boulder commitment is one slice of a much larger footprint. On its official website, Argentum reports 77,000 GPUs contracted for deployment, 344 megawatts of contracted power and 13 global data center sites under exclusivity. Those are company-wide pipeline figures — roughly 34 times the initial 10 megawatts going into Boulder — not measures of the Kyndryl site itself.

The companies intend to broaden the collaboration as Argentum's enterprise AI infrastructure needs evolve. Boulder is the immediate test of that intention: an active GPU deployment that could anchor a broader infrastructure relationship if the expansion plans turn into further commitments.

Source: unite.ai

Share this article:

More from Nathan Brooks

Nathan Brooks

Show full bio

News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.

2,352 articles

Related articles

« Previous articleNext article »