Xeal hopes to raise about $2 billion of debt financing over the next 18 months to build out Laitent, The Information reports. The EV-charging startup’s new network would place compact GPU data centers beside charging infrastructure and sell nearby inference capacity to compute customers running AI applications.
EV-charging sites are wired and permitted for their maximum load, even though Xeal says they typically use less than 10% of it. Laitent’s software would route the remaining power to outdoor pods that fit in a parking space, each holding up to 48 Nvidia GPUs, then combine capacity across nearby properties when a customer needs a larger cluster.
That gives Xeal an installed base for the new service. The company says it can tap more than 200 megawatts of permitted and installed capacity across more than 1,600 properties and plans to deploy more than 100,000 GPUs. Compute customers can size a cluster by GPU count or power and select a metro area, while real-estate partners provide the sites.
The model also changes the economics for property owners. Instead of treating chargers only as an amenity, Laitent says owners can host it as a tenant for fixed rent without upfront capital, or co-invest in the compute layer. Inference revenue is intended to pay for electrical work, chargers and ongoing operations. Xeal’s original charging business already manages payments, electrical loads and revenue sharing for property owners.
The financing plan is not a closed facility: The Information says Xeal hopes to raise the debt over 18 months. Laitent launched on September 18, and its first pod has not yet gone online. Xeal says the deployment with JVM Realty should begin operating by the end of 2026, moving the network from announcement to its first installation.
Published versions remain available when an article is updated or corrected.
- Revision 1 · Initial publication
Initial publication