Blockmate has ordered its first NVIDIA B200 node as customer demand emerges.
Blockmate Ventures Inc (TSX-V:MATE, OTCQB:MATEF) has ordered its first NVIDIA B200 compute node, selected a US data centre to host it, and identified initial customer demand.
The eight-GPU node will be the first AI compute hardware Blockmate owns outright. It is scheduled for installation in the first quarter of 2027, subject to vendor timing, NVIDIA allocation and commissioning.
Blockmate and partner Scannit are already in commercial discussions with prospective customers ahead of deployment.
The company plans to use the node to gather data on utilization, rental pricing, uptime and operating costs before committing additional capital to AI compute.
Chairman Domenic Carosa said the company is now moving from strategy into execution, with progress across each part of the business.
"Our objective remains straightforward: build an AI infrastructure business across power, compute and customers, while deploying capital in a disciplined way as each opportunity develops," Carosa told shareholders.
Blockmate is also working with the local city on planning and permitting for its Wyoming AI data centre. The industrial-area site is under the company's contractual control, and the company is targeting up to about 200 MW of AI and high-performance computing capacity, subject to zoning, engineering and permitting requirements.
In Australia, the company is negotiating several data centre opportunities in power-constrained markets.