ECL Launches 35MW FlexGrid-Powered Data Center in Santa Clara to Meet Rising AI Compute Demand
ECL, a pioneer in sustainable modular data center infrastructure, has announced the development of CSC-1, a new 35-megawatt (MW) data center located in Santa Clara, California. The facility is powered by ECL’s proprietary FlexGrid architecture, a hybrid energy system designed to overcome growing power constraints in high-density artificial intelligence (AI) computing environments.
The CSC-1 site represents a major step forward in how data centers are powered and scaled in one of the most energy-constrained technology hubs in the world. By integrating multiple energy sources—including on-grid electricity, hydrogen fuel cells, and natural gas—the facility is designed to ensure continuous, reliable, and scalable power delivery for AI workloads. This approach directly addresses the increasing pressure on traditional grid infrastructure as demand for AI computing continues to accelerate across the United States.
Meeting Explosive AI Power Demand
The timing of ECL’s announcement reflects a broader industry challenge: rapidly rising electricity demand from AI data centers. According to projections from S&P Global, total U.S. data center electricity consumption is expected to more than double, increasing from 61.8 gigawatts (GW) in 2025 to 134.4 GW by 2030. At the same time, Bloomberg has reported that nearly half of all planned data center projects in the United States this year are expected to face delays or cancellations due to power availability constraints.
This widening gap between demand and grid capacity has created a structural bottleneck for AI infrastructure expansion. Large-scale AI models require enormous computational power, often operating at high rack densities that place significant stress on existing electrical infrastructure. In many cases, utility interconnection delays can stretch into multiple years, making it difficult for operators to deploy new capacity quickly enough to keep up with AI training and inference workloads.
ECL’s CSC-1 facility is designed specifically to address this challenge by reducing reliance on traditional grid timelines and enabling faster deployment through modular, behind-the-meter power generation.
High-Density, Scalable Design
At full buildout, CSC-1 will deliver 35MW of total facility capacity. The project will be deployed in phases, beginning with an initial 2.5MW installation that allows early-stage operations to begin while additional capacity is added over time.
The facility is engineered to support high-density computing environments, with rack power densities ranging from 75 kilowatts (kW) to 270 kW at launch. These density levels are optimized for advanced AI workloads, including large-scale model training, distributed inference systems, and other compute-intensive applications that require substantial and sustained power delivery.
This phased development model allows tenants to begin deploying AI infrastructure within months rather than years. As demand grows, additional modular power blocks can be integrated seamlessly, ensuring that compute capacity scales in parallel with workload requirements rather than being constrained by upfront infrastructure limitations.
FlexGrid: A Multi-Source Energy Architecture
At the core of CSC-1 is ECL’s FlexGrid architecture, a proprietary energy system designed to operate independently of traditional single-source grid dependency. FlexGrid integrates multiple energy inputs, including utility grid power, natural gas generation, and hydrogen fuel cells, into a unified, dynamically managed power system.
This hybrid structure allows the data center to operate in both grid-connected and grid-independent modes, providing resilience against outages, congestion, or interconnection delays. By combining diverse energy sources, FlexGrid ensures continuous uptime while optimizing for efficiency, cost, and emissions reduction.
ECL reports that the FlexGrid system is capable of achieving a sub-1.15 Power Usage Effectiveness (PUE), a metric that measures overall data center energy efficiency. A lower PUE indicates that more of the facility’s power is directed toward computing rather than overhead systems such as cooling and power conversion. This level of efficiency is achieved through a combination of direct-to-chip liquid cooling, air cooling systems, and hydrogen-based energy integration, including water produced as a by-product of hydrogen generation that can be reused in cooling loops.
The result is a system designed not only for high performance but also for reduced environmental impact, particularly in regions like California where water conservation is a critical concern.
Real-Time Infrastructure Intelligence
CSC-1 will also deploy ECL Lightning, a real-time infrastructure management platform that provides granular control over power generation, cooling systems, and rack-level operations. This system enables continuous micro-adjustments to optimize energy distribution and thermal performance across the facility.
By dynamically balancing energy inputs and cooling demand, the platform helps maintain system stability while improving efficiency and reducing waste. It also enhances operational visibility for tenants, allowing them to better manage AI workloads in real time.
Why Santa Clara Matters
Santa Clara sits at the heart of Silicon Valley and remains one of the most strategically important—and power-constrained—data center markets in the United States. The region has become increasingly difficult for new developments due to long utility interconnection queues, regulatory complexity, and limited available grid capacity.
In this environment, power availability has become a defining factor in determining where and how AI infrastructure can be deployed. As a result, behind-the-meter power solutions—systems that generate electricity on-site rather than relying entirely on external utilities—have shifted from experimental approaches to mainstream infrastructure strategies.
ECL’s CSC-1 project reflects this shift. By deploying modular power generation directly at the site, the facility avoids long waiting periods associated with grid upgrades and interconnection approvals. Instead, capacity can be added incrementally, allowing operators to begin production workloads while traditional grid-dependent facilities in the same region remain in development stages.
This model significantly reduces time-to-deployment, a critical factor in the fast-moving AI industry where delays in infrastructure can directly impact competitiveness and innovation cycles.
Water Efficiency and Sustainability Considerations
Sustainability is another central component of ECL’s design philosophy. Data centers are increasingly scrutinized for their energy and water consumption, particularly in drought-prone regions such as California.
At ECL’s pilot MV1 facility, hydrogen is already being used as part of the FlexGrid system. One of the unique advantages of hydrogen-based power generation is that it produces water as a by-product. This water can be captured and reused in the cooling system, reducing or potentially eliminating the need for external freshwater sources.
This closed-loop approach helps address growing concerns about water usage in high-density computing environments and positions CSC-1 as part of a broader shift toward more resource-efficient infrastructure design.
A New Model for AI Infrastructure Deployment
According to ECL co-founder and CEO Yuval Bachar, the CSC-1 project represents a fundamental rethinking of how data center infrastructure is built and scaled.
“A 35MW facility delivered in Santa Clara in under a year would have been unthinkable through traditional grid-connected development,” Bachar said. “Every major AI operator in the Bay Area is facing the same challenge—years-long interconnection queues versus AI deployment needs that are accelerating rapidly. By phasing growth through modular power blocks, ECL aligns infrastructure deployment with actual AI demand rather than forcing customers to overbuild or wait.”
He added that CSC-1 demonstrates how power architecture itself can become an enabling layer for AI expansion rather than a limiting factor
Source Link: https://www.businesswire.com/
