AMD Nets Another AI Success: Millions to Acquire New MI355X

TensorWave has announced the closing of a $350 million Series B funding round that will allow the company to accelerate the expansion of its AI-dedicated infrastructure. The round was led by Magnetar and AMD Ventures, with participation from Maverick Silicon, Nexus Venture Partners, and Western Frontier.

The company, which presents itself as a cloud AI platform based entirely on AMD Instinct GPUs, will use the newfound capital to increase the computational capacity available through the implementation of clusters based on AMD Instinct MI355X accelerators.

The goal is to meet the growing demand for resources for memory-intensive workloads, such as training large language models (LLMs), high-throughput inference, and generative AI applications.

According to TensorWave, the approach based on an open ecosystem and infrastructure optimized for high bandwidth may be particularly interesting for businesses and organizations that need to rapidly scale their AI projects.

Among the clients already operational on the platform are companies specialized in AI such as Fireworks AI and Luma AI, which use TensorWave’s AMD infrastructure to run large-scale generative workloads and production inference systems.

The expansion comes after a completed Series A funding round in May 2025. Since then, TensorWave has built one of the largest AI training clusters based on AMD GPUs in North America, with 8,192 Instinct MI325X accelerators already operational.

On the infrastructure front, TensorWave also claims to have secured over 2 gigawatts of long-term capacity for its data centers, resources that are expected to support the growth of demand from enterprise clients, research centers, and AI-native companies. The company also plans to expand its workforce at its Las Vegas headquarters, with new hires in engineering, infrastructure operations, sales, and customer support.

For AMD, this operation represents another sign of the growing interest in the Instinct family in the AI market. The MI355X is indeed one of the company's main offerings for AI data centers and precedes the arrival of the future Instinct MI455X GPUs, designed to further enhance performance and improve total cost of ownership for infrastructures. These solutions will be called to compete with the upcoming NVIDIA platforms destined for next-generation data centers.