Seguente offers COLDWARE, a passive two‑phase direct‑to‑chip liquid‑cooling system that eliminates pumps and water loops by using a dielectric coolant that boils and returns via buoyancy‑driven vapor flow. The technology enables rack densities of 250 kW+ (and over 1 MW per cluster) with PUE as low as 1.03, delivering up to 80% total‑cost‑of‑ownership reduction and near‑zero maintenance for AI, HPC, and cloud data centers. Its modular, waterless design can be retrofitted into existing facilities or deployed in new builds, providing high‑density, energy‑efficient cooling for hyperscale and enterprise operators.
Funding
Funding not disclosed
Founders
Product
Problem
Data centers using traditional air‑cooled or pumped liquid cooling systems face high energy consumption, complex maintenance, and significant water usage, which limit rack density and increase total cost of ownership.
Solution
Seguente’s COLDWARE technology provides passive, two‑phase direct‑to‑chip cooling that eliminates pumps and water loops. Heat from CPUs and GPUs boils a dielectric coolant in cold plates, creating a buoyancy‑driven vapor flow that returns to a condenser for condensation, creating a closed‑loop system with no external power for circulation. This enables rack densities of 250 kW+ per rack and over 1 MW per cluster while achieving PUE as low as 1.03. Because the system is waterless and uses stable, non‑conductive coolants, it reduces operational expenses, maintenance effort, and environmental impact. The modular design can be retrofitted into existing facilities or deployed in new builds, offering scalable performance for AI, HPC, and cloud workloads.
Target Audience
Primary customers are hyperscale cloud providers, AI and high‑performance computing operators, and enterprise data centers seeking higher rack density, lower energy costs, and sustainable cooling solutions.
Features
- Passive two‑phase heat transfer using dielectric coolant eliminates pumps and reduces energy use
- Waterless operation removes the need for chillers and minimizes water consumption
- Self‑regulating closed‑loop system provides near‑zero maintenance and high reliability
- Scalable architecture delivers 250 kW+ per rack and 1 MW+ per cluster for high‑density AI/HPC workloads
- Modular rack and cluster units integrate with legacy infrastructure and support flexible deployment
- Low‑temperature operation improves component lifespan and enables higher compute performance
- Compatible with standard server hardware, allowing direct retrofits of existing CPUs/GPUs