Crusoe Energy utilizes stranded energy from oil flares to power their AI data centers, reducing carbon emissions and energy costs.
Crusoe's cloud infrastructure is designed to prioritize energy-intensive computing workloads with a focus on energy efficiency and high-performance networking.
Crusoe's digital flaring mitigation solution captures waste gas from oil wells and converts it into energy, reducing emissions by 70% and extending the climate runway.
Deep dives
Digital infrastructure tackling energy-intensive computing problems
Crusoe is a cloud infrastructure provider that focuses on aligning the future of computing with the future of the climate. They prioritize energy-intensive computing applications and aim to make the impact scalable and cost-efficient. By building data centers next to oil fields with active oil flares, Crusoe takes advantage of wasted energy and uses it to power their AI data centers. The goal is to reduce energy costs and carbon emissions by utilizing these stranded energy resources.
Crusoe's infrastructure is designed to cater to energy-intensive computing workloads. They focus on compute, networking, and storage, with a strong emphasis on energy efficiency. Their infrastructure is built for high power density racks, enabling efficient heat management and high-performance networking. By using technologies like remote direct memory access, they optimize server-to-server communication. Their compute infrastructure is delivered through a KVM-based virtualization stack, providing customers with a compute-first platform.
Reducing greenhouse gas emissions from flaring
Crusoe's digital flaring mitigation solution addresses the issue of flaring in the oil industry. They capture waste gas from oil wells that would otherwise be burned and convert it into energy to power their data centers. By eliminating the greenhouse gas footprint of flaring, Crusoe reduces emissions by about 70% while providing a beneficial use for the stranded energy resource. This approach helps extend the climate runway by reducing emissions from legacy industrial sources.
Optimizing renewable energy facilities with digital infrastructure
Crusoe also partners with renewable energy producers to optimize their facilities. They bring demand to sites with stranded or negatively priced renewable power, helping these facilities monetize their excess energy and eliminate negative pricing risk. By powering their data centers with onsite renewable carbon-free power, Crusoe aims to achieve a net-zero carbon footprint. This approach aligns with their mission of aligning the future of computing with the future of the climate.
Thinking like a mountaineer
Crusoe Energy Systems has a company value of 'thinking like a mountaineer' which prioritizes safety orientation and being prepared for potential challenges.
Unique capital structure
Crusoe Energy Systems has raised $500 million in venture equity funding from investors such as Founders Fund, Bing Capital Ventures, and Valor Equity Partners. They have also employed unique financing methods, including asset-backed financing and project financing facilities, to fund their physical infrastructure and equipment.
We sit down with Crusoe Energy CEO Chase Lochmiller to talk about the two “hard to imagine” tasks they’ve undertaken: 1) building a new AI cloud infrastructure provider from scratch, and 2) colocating and powering it with stranded energy from some of the harshest and most remote locations on earth.
Crusoe’s cloud of course has to compete with (and in many cases exceed) the price/performance curves of cloud incumbents like AWS, Azure and Google in processing AI workloads. And the way it does so is by building data centers literally on top of oil flares (and other wasted energy sources) that otherwise comprise multiple percentage points of annual global greenhouse gas emissions. In other words — methane that previously just got lit on fire is now powering your favorite AI startup’s training workloads!
We cover what it actually takes to build and operate a public cloud, the latest Nvidia networking and server innovations and what they mean for GPU data centers, and how to set up a company to pursue something “hard” like this across the team, operations and capital raising fronts. Tune in!