TensorWave has announced a $100 million Series A funding round to scale the deployment of what it claims is the world’s largest liquid-cooled AMD GPU cluster. Featuring 8,192 AMD Instinct MI325X GPUs, the purpose-built infrastructure is optimized for AI training, fine-tuning, and inference workloads at massive scale. The round was co-led by Magnetar and AMD Ventures, with participation from Prosperity7, Maverick Silicon, and Nexus Venture Partners.
The company has gone all-in on AMD, citing the MI325X’s 256GB of HBM3e memory, open ROCm support, and unmatched compute density as core enablers of its high-performance cloud. Direct liquid cooling eliminates thermal bottlenecks and supports higher GPU density and energy efficiency, enabling consistently high throughput without thermal throttling. “We engineered our system from the ground up to make high-density, high-performance clusters viable—and liquid cooling is the unlock,” said Piotr Tomasik, President & COO of TensorWave.
“We’re not here to offer another cloud—we’re here to build the one that AI actually needs.”
— Darrick Horton, CEO, TensorWave
- TensorWave raises $100M Series A to build largest liquid-cooled AMD GPU cloud.
- Cluster includes 8,192 MI325X GPUs, each with 256GB of HBM3e memory.
- Company is fully AMD-native, optimized for ROCm and open-source model development.
- Direct liquid cooling architecture enables higher density and sustained performance.
- Investors include Magnetar, AMD Ventures, Prosperity7, Maverick Silicon, and Nexus Venture Partners.
TensorWave, headquartered in Las Vegas, Nevada, is an AI and HPC cloud infrastructure company founded in 2023 by Darrick Horton (CEO), Jeff Tatarchuk (Chief Growth Officer), and Piotr Tomasik (President). The founding team brings a wealth of experience from previous ventures, including VMAccel, VaultMiner Technologies, and Influential, providing a strong foundation in cloud computing, high-performance infrastructure, and technology startups. Their shared vision to address the challenges of GPU availability and performance bottlenecks in AI workloads led to the creation of TensorWave, focusing exclusively on AMD’s Instinct™ Series GPUs to deliver high-bandwidth, memory-optimized infrastructure.