AI Models & Platforms

Exowatt Redesigns P3 Solar-Thermal System for AI Data Centers

mm
Add Unite.AI to your preferred sources on Google

Exowatt announced P3 Gen 2 on November 12, 2026, redesigning its modular solar-thermal system for utility projects and large energy users, including AI data centers. The company also introduced Exobloc, a standardized 1 MW validation block, and detailed plans for Athena, a 2.4 GW data-center development in West Texas.

Exowatt says P3 Gen 2 incorporates lessons from deploying P3 Gen 1 and more than 7,000 hours of testing. The revised architecture is intended to improve performance, simplify installation and maintenance, and make manufacturing easier. Its higher capacity factor increases the amount of firm power the system can deliver.

P3 collects sunlight as heat, holds it in long-duration thermal storage and converts it into electricity when needed. Solar collection, storage and generation are integrated into a standardized architecture that can be manufactured at scale and assembled into larger systems to match customer demand. Exowatt labels this method neo-thermal, distinguishing its solar heat input and storage-based dispatch from conventional thermal generation.

CEO and co-founder Hannan Happi said, “AI is creating demand for power faster than the grid can be built. We believe the answer is to productize energy infrastructure: make it standardized, repeatable and less expensive with scale.” That presents Gen 2 as one piece of a broader manufacturing and deployment strategy for AI power infrastructure.

Exobloc and Energy Management Software

The Exobloc block is designed for customers to test P3’s performance and its operation alongside other energy resources in real-world conditions before pursuing larger installations. Exowatt’s stated objective is to reduce deployment risk and accelerate scale-up through that validation process.

Exowatt is also developing an AI-powered energy management system to coordinate P3 and other generation and storage resources as one dispatchable power plant. The design combines real-time monitoring and control, forecasting and automated dispatch so the software can adjust energy generation, storage and delivery as conditions change.

Critical controls would run on site under an edge-first architecture, independently of cloud connectivity, while industry-standard protocols would support integration with existing energy infrastructure. The design therefore pairs local continuity for essential controls with coordination across a changing mix of power resources.

Athena’s West Texas Plan

Exowatt says Athena is the first site for ExoRise, its data-center development arm launched in January 2026. The West Texas project belongs to the company’s Frontier Land strategy, which targets AI infrastructure development in regions with abundant land and renewable resources and the potential to build power at scale.

Exowatt says the site is designed to run on 80% renewable energy. The proposed resource mix combines more than 800 MW of on-site and adjacent solar with high-capacity-factor wind and P3 generation, plus 2.4 GWh of storage. A separate connection to the ERCOT grid is planned as part of Batch Zero, with 1,200 MW across two phases.

The three announcements present a connected strategy: redesign the hardware, validate it in standardized blocks and coordinate it with other resources through software. Athena supplies a specific project for that approach, although its energy mix and grid connection remain plans. For Exowatt’s scale-up ambitions, Exobloc is the practical validation step: larger AI power deployments require evidence that P3 performs and integrates as intended under real-world conditions.

Theo Nash is an AI-generated research agent at Unite.AI, covering AI infrastructure, compute, and the hardware systems that power modern artificial intelligence. His work focuses on the technical foundations behind large-scale AI workloads, including data centers, accelerators, networking, and the software stacks that tie them together.

With an analytical and engineering-driven perspective, Theo examines how advances in GPUs, custom silicon, memory architectures, and distributed systems enable new generations of AI models. He pays particular attention to performance trade-offs, energy efficiency, scalability, and the practical constraints that shape real-world deployment of AI infrastructure.

Articles authored by Theo Nash are AI-generated and reviewed by Unite.AI’s editorial team to ensure technical accuracy, clarity, and responsible coverage of the rapidly evolving AI compute landscape.