A startup betting against Nvidia's dominance in AI data centers just got a major vote of confidence. TensorWave has raised $350 million in fresh funding, pushing its valuation to $1.55 billion, according to the Wall Street Journal. The company plans to use the capital to expand its data centers — filled not with Nvidia's sought-after GPUs, but with chips from AMD, which is also an investor in the company.
TensorWave isn't alone in challenging the status quo. D-Matrix, a Microsoft-backed startup founded by Indian-American entrepreneurs, is also taking direct aim at Nvidia with its Corsair AI inference chip. According to CNBC TV18, D-Matrix claims its technology can deliver faster, cheaper, and more energy-efficient AI workloads, and the company is working to build an entirely new data center ecosystem around its hardware.
The broader picture: Nvidia controls an estimated dominant share of the market for AI accelerator chips, making its hardware the default choice for cloud providers and startups racing to build AI infrastructure. That near-monopoly position has drawn intense scrutiny — and inspired a wave of well-funded challengers.
AMD's role as both a chip supplier and an investor in TensorWave signals that the semiconductor giant sees real business in helping data center operators diversify away from a single vendor. For AI companies and cloud customers, more competition means potential relief from the long wait times and high costs that have defined the Nvidia chip market.
If even a fraction of new data center spending shifts toward AMD or homegrown alternatives, it could reshape the economics of the AI infrastructure boom that has defined the tech industry over the past two years.