Torus announces $500 million in growth capital

Torus said the growth capital will support GigaOne’s expansion and deployments of Torus Reserve and Foundry across the United States.[1] Magnetar led the round; ICONIQ, Origin Ventures, Ridgeline Partners and Trinity Capital also participated.[1] Torus says Foundry is designed, engineered and manufactured in Utah by a team of nearly 400 employees at GigaOne.[1] The company says GigaOne opened its first production line in early 2026 and opened a new line every 21 days during 2026.[1]

Foundry packages power, storage, cooling, networking and compute into a prefabricated system.[1] It is installed behind the meter, using available electricity at individual transformers, and every system is paired with Torus Reserve, the company’s energy-storage and management product.[1] Torus says Foundry provides backup power during grid outages.[1] The company says the smallest system is roughly the size of a small cargo van, and that site preparation takes about a week, followed by one to two days for installation and commissioning.[1] Torus also offers a powered shell with power and an enclosure but no GPUs.[1]

Torus says Foundry provides 1 to 20 megawatts of inference capacity per site and can go from contract to running AI workloads in 12 to 16 weeks.[1] The company named Hydra Host and another, unnamed AI cloud provider as its first customers.[1] It says Foundry has thousands of GPUs deployed and plans to exceed 47,000 GPUs in Foundry units by the end of 2027.[1] Torus says its production line is online and can produce 500 megawatts of IT load, with plans to expand to 1 gigawatt.[1] It cites McKinsey projections of 35% annual growth in data-center demand for AI inference through 2030, versus 22% for training.[1]