TensorWave lands $350 million Series B as it builds more AMD-based AI capacity at a $1.55 billion valuation

TensorWave, a cloud company focused on running artificial intelligence workloads entirely on Advanced Micro Devices hardware, said it raised $350 million in a Series B round that values the business at $1.55 billion. The announcement, made June 10, also points to an emphasis on scaling infrastructure for large, memory-intensive AI tasks, including training large language models and running high-throughput inference for generative applications.

The round was co-led by Magnetar and AMD Ventures, according to the post circulating through financial news channels. Other investors listed as continuing participants include Maverick Silicon, Nexus Venture Partners, and Western Frontier, reflecting a broader push by both strategic and financial backers to fund alternative approaches to AI compute supply.

TensorWave describes itself as an “all-AMD” AI cloud, targeting what it calls memory-intensive workloads. In its growth plan, the company said the new capital will help expand the footprint of its global AI infrastructure. It also highlighted deployments of next-generation AMD Instinct MI355X GPU clusters, which it described as suited for memory-heavy use cases such as large language model training, high-throughput inference, and generative AI applications.

The company’s positioning is aimed at shifting AI capacity from experimentation to production, particularly as model sizes continue to rise. TensorWave’s CEO and co-founder, Darrick Horton, said the next phase of AI will be defined by which companies can access enough compute to move from tests to real deployments, arguing enterprises need infrastructure with the memory capacity, performance, and flexibility to scale without being locked into a single ecosystem.

TensorWave also said it is seeing adoption from next-generation AI companies that use its AMD-based infrastructure for large-scale generative AI workloads and production inference systems. It cited Fireworks AI and Luma AI as examples of customers building on its platform, though it did not provide additional detail on the size of these deployments or the commercial terms behind them.

From a technology and market standpoint, TensorWave’s funding arrives as competition in AI infrastructure continues to broaden beyond a single chip supplier. AMD has been actively positioning its data center accelerators and software ecosystem for training and inference, while startups and cloud operators have increasingly sought to differentiate based on cost, performance per watt, access, and supply resilience. TensorWave’s emphasis on “open-ecosystem” infrastructure and memory capacity reflects one way providers are trying to compete for enterprise AI budgets.

Still, the announcement does not spell out several items investors and customers typically want when evaluating an infrastructure platform, including where the expansion will occur geographically, the expected timeline for rolling out additional capacity, or how performance and availability compare with other hyperscale and managed-GPU offerings. The disclosure also does not specify whether the Series B will support new data center leases, build-outs, or partnerships, nor does it provide revenue or customer retention metrics.

Looking ahead, the immediate next question is how quickly TensorWave can translate the new funding into deployed GPU capacity and demonstrable workload performance for memory-bound training and inference. Observers may also watch whether AMD’s involvement via AMD Ventures deepens into further commercialization support, and whether other investors join follow-on rounds as demand for AI compute continues to strain capacity across the industry.