TensorWave Closes 350,000,000 Series B Financing Led by Magnetar and AMD Ventures to Deploy Next Gen AMD Clusters for Global Scale
Cloud provider specialized in hardware for intensive compute workloads, TensorWave, has closed a $350,000,000 Series B financing round. The financing round, as detailed in TensorWave's official press release, was co led by Magnetar and AMD Ventures. Additional investment was received from Maverick Silicon, Nexus Venture Partners and Western Frontier. This capital will directly enable the company to scale its physical infrastructure footprint globally.
The capital will be immediately put toward procuring and deploying the next gen AMD Instinct MI355X GPU clusters. These systems are designed to accommodate high performance, memory heavy enterprise workloads, such as the training of large language models, the generation of graphics and high throughput inference. By using an open hardware ecosystem, the startup gives corporate clients an alternative to the capacity constrained, vertically integrated supply chain currently limiting availability.
The enterprise space has demonstrated robust initial interest in this type of hardware setup. Generative AI firms like Fireworks AI and Luma AI have already deployed the company's large scale production workloads onto the infrastructure, which is able to provide them the required high bandwidth capacity to scale without encountering hardware shortages like others in the market.
After swift growth post Series A funding round, the company is currently running a large AMD cluster in North America, operating 8192 AMD Instinct MI325X GPUs online. It has already secured 2+ GW of long term power for the procurement and deployment of further MI355X systems into multiple new data centers. A percentage of this funding round will be channeled towards expanding its personnel in its headquarters in Las Vegas to include engineers, salespeople, operators and support roles to provide global support around the clock to customers worldwide.
"The current phase of industry growth depends entirely on providing compute capacity where it's needed," explained Darrick Horton, chief executive officer and co founder of the startup. "Businesses need a level of scalable compute capacity that they can utilize without getting locked into a single software stack."
"The next evolution of AI will be decided by those who can obtain the compute necessary to scale their efforts from research to production"
Horton continued to say that, as modern systems have grown to extreme sizes, memory capacity and hardware flexibility are now critical.
"As models grow larger and workloads more computationally demanding, enterprises need the infrastructure that supports massive memory capacity, high performance, and extreme flexibility in order to scale beyond the limitations imposed by proprietary ecosystems."
This injection of capital puts the company in a prime position to be an alternative provider of specialized hardware for intensive enterprise workloads, leveraging AMD's hardware to bypass the current supply bottleneck in silicon.
