Torus, a Salt Lake City energy-storage startup, said it raised $500 million in a Series C round led by Magnetar to help launch a shipping container-sized product that combines batteries with artificial-intelligence computing, Bloomberg reported on 11 November, citing a Torus statement.
The round is 2.5 times the $200 million Magnetar invested in Torus in September 2025, by our arithmetic. Iconiq Capital, Origin Ventures and Trinity Capital also took part, and the AI infrastructure company Hydra Host is among the customers, Torus said.
Business Pill · Grid Cost Allocation
An evergreen Business Engineer explainer on who pays for the grid capacity AI demands. It teaches the general idea only and does not discuss Torus.
The key insight: As we read it, the money follows the product, not the flywheel: Torus raised 2.5 times its 2025 Magnetar cheque, by our arithmetic, to manufacture a container that puts AI computing next to its batteries.
What the $500 Million Is For
The funds will support manufacturing of the new product, called Foundry, according to the statement as reported by Bloomberg. Foundry pairs batteries with AI computing in one shipping container-sized unit.
That is a shift in what Torus sells. In September 2025 it described itself as “a full-stack energy platform offering storage, management, security, and generation solutions”, and its systems as small, inertia-based hybrids that combine mechanical flywheels with batteries.


Update (11 Nov): Torus’s own release gives the specifications. Torus has published its announcement of Foundry and the $500 million round. It says Foundry is “prefabricated and modular, at 1–20 MW per site, with no building construction required”, and goes “from contract to running AI workloads in 12–16 weeks”; site prep “takes about a week, and installation and commissioning take one to two days”. Torus says Foundry “has thousands of GPUs deployed today, with plans to have more than 47,000 GPUs deployed in Foundry units by the end of 2027”, and that the Foundry production line “will be capable of producing 500 MW of IT load, with plans to expand to 1GW in the near future”. That 500 MW equals 25 sites at the 20 MW top size, by our arithmetic. The release names Hydra Host and another AI cloud provider as first customers and adds Ridgeline Partners to the round. Torus’s news page now lists $767M raised, $500 million more than the $267 million cited above, by our arithmetic.
The Torus Base the Round Builds On
Torus’s website lists 400 employees, a 545,000 square foot flagship factory in Utah called GigaOne, 12 to 16 weeks from contract to operation, 99.9% fleet uptime and $267 million raised. The new round is about 1.9 times that $267 million, by our arithmetic.
Its 2025 release said Torus was expanding its partnership with the utility PacifiCorp through a memorandum of understanding covering up to 500 megawatts of demand response capacity, after a 70 megawatt agreement that was filled within six months.
Power is the constraint other AI builders are also working around: Crusoe cut its power gear wait by 40%, and SpaceX has pitched a 6GW-plus supercomputer campus in Arizona.
The Structural Read
As we read the two documents together, Torus is moving from selling grid storage to utilities and factories toward selling a box that also runs AI workloads, with an AI infrastructure company, Hydra Host, named as a customer.
As we read the figures, the $500 million round is about 1.9 times the $267 million Torus’s website lists as raised, by our arithmetic, so the company’s funding more than doubles in one round.
Dave Snyderman, Magnetar, in Torus’s September 2025 release
“Torus is at the forefront of redefining energy infrastructure, and this investment will allow the company to accelerate its growth and bring this new utility model to market at scale.”
Three Implications
FOR AI BUILDERS Foundry pairs batteries and AI computing in one shipping container-sized unit, according to the statement as reported by Bloomberg; its specifications are not in the reports we read.
FOR UTILITIES Torus’s 2025 release described a memorandum of understanding with PacifiCorp covering up to 500 megawatts of demand response capacity, after a 70 megawatt agreement.
FOR INVESTORS Magnetar led both the $200 million investment of September 2025 and the $500 million Series C, per the release and Bloomberg.
The Business Engineer Lens
This story maps onto the Business Engineer framework The State of AI Data Centers.
The framework argues: “The structural bottleneck isn’t computing, it’s electrons.”
As we read it, Foundry is a bet on that bottleneck: storage and compute shipped in the same container. The framework predates the round and does not discuss Torus.
What Is Not Established
The reports we read give no valuation, no Foundry specifications such as storage capacity or the number of chips per unit, no price and no delivery dates. We could not read Torus’s own statement; it was not on its website when we checked, so the round’s details rest on Bloomberg’s account of it.
Bloomberg names Hydra Host as one customer; the terms and the size of that order are not in the reports we read. We did not contact Torus.
The Bottom Line
Torus says it raised $500 million, led by Magnetar, to make Foundry, a container that combines batteries with AI computing; the round is 2.5 times Magnetar’s $200 million investment of September 2025, by our arithmetic.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. The round, its investors, Foundry and Hydra Host are from Bloomberg’s report of 11 November 2026 on a Torus statement; we read the report’s opening and could not find the statement itself on Torus’s website or the newswires when we checked. The $200 million Magnetar investment, the PacifiCorp figures and Torus’s 2025 self-description are from its release of 9 September 2025, read in full. The headcount, factory size and $267 million raised are from its website. We did not contact Torus.
Sources: Bloomberg, 11 Nov 2026 · Torus release, 9 Sep 2025 · Torus website









