TensorWave, a startup providing a cloud platform for AI applications that runs on chips designed by an Nvidia rival, has raised $43 million in funding.
The Las Vegas startup, which launched in 2023, runs data centers that deploy "thousands" of GPUs. Customers can access them via the cloud to run AI workloads.
Darrick Horton, the cofounder and CEO of TensorWave, told BI that it only uses hardware provided by semiconductor company AMD.
The team saw a problem in the AI compute space wherein Nvidia had a "monopoly" and companies lacked a "viable alternative," cofounder and chief operating officer Piotr Tomasik said.
The startup's pitch deck, shared exclusively with BI, states that "buyers desperately need an alternative for AI compute infrastructure."
TensorWave makes its money in two key ways. "We used a fairly standard business model for neoclouds, which is charged by the hour, so it's infrastructure as a service," Tomasik said.
"We also tend to have reserve time contracts. So that means someone might rent eight GPUs from us for an entire year, and they pay us for that year, regardless of whether they use it for 10 hours or 24 hours, seven days a week," he told BI.
The funding came in the form of a simple agreement for future equity, or SAFE, a financial instrument that means investors don't get a stake until certain criteria are met.
Nexus VP led the $43 million investment, which the startup claimed is the largest SAFE round in Nevada's history. Additional funding came from Maverick Capital, Translink Capital, Javelin Venture Partners, StartupNV, and AMD Ventures.
While fundraising, the team said it observed growing momentum in the GPU cloud space and increased investor appetite.
With the cash injection, TensorWave plans to increase its head count and deploy more hardware.
Check out the 9-slide pitch deck TensorWave used to secure the fresh funding.
