Crusoe and Redwood Scale AI via Recycled Batteries
- •Crusoe and Redwood Materials expand partnership, scaling AI infrastructure density sevenfold in Nevada
- •New 12MW microgrid utilizes second-life electric vehicle batteries to power high-performance compute clusters
- •System achieved 99.9% total uptime by combining recycled battery storage with solar and grid backup
The explosive growth of artificial intelligence has hit a physical wall: the massive energy requirements of high-performance data centers. Crusoe and Redwood Materials are addressing this bottleneck by scaling their joint operations in Nevada to seven times their original density. By deploying 24 modular data centers—which are prefabricated units that can be set up in months rather than years—the partnership bypasses the long lead times of traditional construction while maintaining the high-performance computing (HPC) power necessary for modern AI workloads.
What makes this expansion unique is the reliance on "second-life" batteries. When electric vehicle batteries no longer meet the rigorous standards for automotive use, they still retain significant capacity for stationary storage. Redwood Materials repurposes these batteries into a 12-megawatt microgrid, managed by their proprietary Pack Manager technology. This system orchestrates energy flow from on-site solar panels and the stored battery power to ensure a steady, reliable stream of electricity for Crusoe’s hardware, which handles the intense math required for AI training and inference.
The results from their initial seven-month pilot are promising, boasting 99.2% operational availability on renewable sources alone. By using the traditional power grid as a secondary fail-safe, the partnership achieved a total uptime of 99.9%. This "AI factory" model suggests a sustainable path forward, transforming the waste of the transition to electric vehicles into the fuel for the AI revolution. As compute demand fluctuates, the modularity of this infrastructure allows providers to scale capacity quickly, ensuring that the supply chain for intelligence stays ahead of the curve.