Xeal Laitent Taps Idle EV Charging for AI

Xeal has launched Laitent, an edge inference network that uses idle EV charging capacity. The company plans 100,000 NVIDIA GPUs across 1,600-plus properties, with the first Pod online by the end of 2026.

Xeal has launched Laitent, an edge inference compute network that the company says will draw on more than 200 MW of permitted, installed electrical infrastructure across 1,600-plus EV charging properties. The EV charging provider, which reports one of the five largest Level 2 charging footprints in the U.S., plans to deploy more than 100,000 NVIDIA GPUs alongside its chargers. Xeal describes Laitent as the first network of its kind, and the first Laitent Pod is scheduled to come online with property partner JVM Realty by the end of 2026.

Highlights

  • Capacity: More than 200 MW of permitted, installed electrical infrastructure across 1,600-plus properties, according to Xeal.
  • Hardware: More than 100,000 NVIDIA GPUs planned, with up to 48 GPUs in each Laitent Pod.
  • Utilization: Xeal says EV charging sites typically run at less than 10% of permitted capacity, leaving the remaining 90% for compute.
  • Timing: The first Laitent Pod goes online with JVM Realty by the end of 2026.

How Does Laitent Use Idle EV Charging Capacity?

EV charging sites must be permitted for a maximum charging load. According to Xeal, those sites typically operate at less than 10% of that permitted capacity, and Laitent uses the remaining 90% for compute.

The company says this approach sidesteps the most difficult hurdle in bringing new compute online: grid interconnect. Xeal’s dynamic power orchestration software routes idle energy capacity within each Pod. The company says the software can activate anything from a single slice of a single GPU to a distributed network spanning a metropolitan area.

“If you need inference compute today and can’t wait for a new data center to come online, Laitent can accelerate your timeline from years to months,” said Nikhil Bharadwaj, co-founder and CEO of Xeal. “Laitent is the fastest way to bring new compute online, tapping already permitted and grid-connected, yet idle energy capacity, located right where it’s needed for ultra-low latency, with no water hookup requirements.”

Xeal places the Pods in metro areas, a location strategy Laitent calls the Metro Edge. The company reports sub-20 ms latency from these sites.

Laitent Pod Specifications

Each Pod occupies about one parking space. Xeal lists the following features:

  • Compute: Up to 48 NVIDIA Hopper or Blackwell Ultra GPUs
  • Cooling: Self-contained, independent cooling with no water hookup required
  • Noise: Less than 65 dB, which the company compares to a washing machine
  • Enclosure: NEMA 4, built for outdoor installation
  • Installation: Installed and running in hours, according to Xeal
  • Power costs: All power bills covered as a tenant passthrough
  • Site owner terms: Fixed rental or ancillary revenue

Xeal also cites potential savings for EV charging customers.

NVIDIA, Rafay Systems and Spectrum Business Partnerships

Xeal is a member of NVIDIA Inception, NVIDIA’s startup program. The company has also secured partnerships with Rafay Systems for AI infrastructure orchestration and Spectrum Business for dedicated enterprise-grade fiber serving inference workloads. Dozens of real estate and property managers have signed on, along with an unnamed Tier 1 inference provider for up to 5 MW of compute.

“As AI adoption grows, infrastructure must become more distributed, flexible and energy efficient—placing the right workloads where power and capacity are available,” said Marc Spieler, Senior Managing Director for the Global Energy Industry at NVIDIA. “By pairing NVIDIA accelerated computing with Xeal’s existing grid-connected EV charging footprint, Laitent offers an innovative way to bring low-latency inference closer to users while making better use of infrastructure that has already been built.”

“The industry has been asking how to bring inference closer to users without years of data center buildout — Laitent is a real answer,” said Haseeb Budhani, CEO of Rafay Systems. “Rafay’s orchestration layer turns Xeal’s distributed EV charging sites into a production-grade, GPU-ready compute fabric, so operators get hyperscaler-level management without hyperscaler-level infrastructure. This is exactly the kind of distributed, energy-aware AI infrastructure the market needs next.”

What Does Laitent Mean for Property Owners?

Xeal says Laitent gives property owners ancillary revenue and can add as much as $1 million in property value for little-to-no upfront investment. JVM Realty, host of the first Pod, has worked with Xeal since 2023.

“With Laitent, Xeal is creating new ancillary revenue streams from infrastructure already in place across our communities,” said Steve Boyack, Chief Operating Officer of JVM Realty. “For property owners, the combination of a new income stream with little-to-no upfront investment creates meaningful financial upside.”

Expansion Beyond 200 MW

Beyond the initial 200 MW of installed charging capacity, Xeal plans to unlock more than 1 GW of existing headroom across real estate and EV charging deployments. The company reports building more than 1,600 EV charging sites over the last seven years. Xeal says its team has sited and installed energy infrastructure across more than 160 U.S. metro areas.

The EV Report
The EV Report Staff

The EV Report is the trade publication of record for vehicle electrification. Published by Hagman Media and edited by founder Brian Hagman, it covers battery electric vehicles, plug-in hybrids, hydrogen fuel cell vehicles, charging infrastructure, and battery technology for an audience of automotive engineers, fleet managers, and clean-mobility investors.