TensorWave Lands $350M Series B At $1.55B To Scale All-AMD AI Cloud

TensorWave's $350 million Series B, co-led by Magnetar and AMD Ventures, gives the AMD-only AI cloud the firepower to chase Nvidia in inference and training.

TensorWave Lands $350M Series B At $1.55B To Scale All-AMD AI Cloud

Las Vegas-based AI cloud startup TensorWave has closed a $350 million Series B round at a $1.55 billion valuation, locking in fresh capital to expand the only major US data-center business that has built its stack entirely on AMD silicon. The round, announced on June 10, was co-led by Magnetar and AMD Ventures with participation from Maverick Silicon, Nexus Venture Partners, and Western Frontier.

An AMD-only bet against Nvidia's monopoly

While almost every AI infrastructure operator is fighting for Nvidia's H-series and Blackwell GPUs, TensorWave runs three operational data centers in Arizona, Florida, and Pennsylvania filled exclusively with AMD Instinct MI325X accelerators. The new capital will fund deployments of the next-generation MI355X across multiple regions in North America, plus expansion into 2 gigawatts of leased capacity.

Stacks of GPU servers in an AI data center

A backstop for AMD's ROCm push

TensorWave has become a key co-development partner for AMD's ROCm software stack, helping smooth the rough edges that have historically slowed adoption versus Nvidia's CUDA ecosystem. Chief Executive Darrick Horton told the Wall Street Journal that ROCm is now "pretty much plug-and-play" for inference workloads, and AMD is increasingly positioning its GPUs as a cost-and-power-efficient alternative for production serving rather than purely for frontier training.

Riding the alt-compute wave

The round triples TensorWave's valuation in roughly 12 months and underscores investor appetite for compute capacity that isn't tied to Nvidia's allocation queues. Public-market peer Cerebras Systems went public last month with its dinner-plate inference chips, and the broader build-out across hyperscalers has pushed projected 2026 AI capex past $700 billion. TensorWave is using the new funding to grow headcount from roughly 160 to 300-400 employees and to reach 100,000 GPUs deployed within a year.

For more context, see our coverage of Apollo's $35B Broadcom AI XPV platform funding Anthropic compute, Applied Digital's $5.2B Delta Forge 2 take-or-pay lease, and NAVER's gigawatt-scale NVIDIA DSX AI factory project.

Reporting based on coverage from SiliconANGLE, BusinessWire, and the Las Vegas Review-Journal.

Category: Funding & Investments

Tags: AI Models venture capital funding Series B Funding AI data centers

Related Articles