Hyperscale Power designs solid‑state transformers that convert medium‑voltage AC directly to 800 V DC for AI data centers and other high‑density loads. Their SSTs combine SiC and GaN semiconductors with advanced magnetics to deliver up to 3.3 MW at 98.5 % efficiency in a compact, modular form factor. This enables data‑center operators to increase power density while reducing footprint, losses, and infrastructure complexity.
Funding
Funding not disclosed
Founders
Product
Problem
High-density computing environments, particularly those supporting AI and High-Performance Computing (HPC) workloads, generate significant thermal loads that traditional air-cooling methods struggle to manage effectively. Inadequate thermal management leads to reduced component lifespan, performance throttling, and increased energy expenditure due to inefficient cooling operations.
Solution
Hyperscale Power delivers advanced liquid cooling solutions engineered for the demanding thermal requirements of modern data centers. Their technology facilitates direct liquid cooling (DLC) or immersion cooling strategies, enabling superior heat dissipation compared to air-based systems. This approach allows for higher compute densities and sustained peak performance from hardware. By optimizing thermal efficiency, Hyperscale Power's solutions reduce overall data center PUE (Power Usage Effectiveness) and enhance operational reliability, ensuring consistent uptime for critical workloads.
Target Audience
The primary customers are data center operators, colocation providers, and enterprises deploying AI/HPC infrastructure that require robust and efficient thermal management solutions.
Features
- Direct-to-chip liquid cooling manifolds for CPU and GPU thermal management
- Immersion cooling tanks utilizing dielectric fluids for complete component submersion
- Integrated pump and heat exchanger modules for closed-loop liquid circulation
- Advanced sensor arrays for real-time temperature and flow monitoring
- Modular design for scalable deployment in existing data center infrastructure
- Energy-efficient heat rejection systems to minimize operational power draw