Issue #14: Crusoe's Flare Gas Profile

Flared gas is a problem.

Oil and gas operations vent billions of cubic feet of methane into the atmosphere each year. This isn’t just wasteful. It’s one of the most climate-damaging emissions sources globally.

Yet, where others saw a problem, one company saw an opportunity.

Instead of letting that gas burn uselessly into the sky, why not convert it into power for AI and high-performance computing?

They built a system that captures flared gas, converts it into electricity, and powers energy-efficient data centres. What started as a Bitcoin mining venture evolved into a full-fledged AI cloud platform, with compute clusters now running NVIDIA’s most advanced GPUs. And now, while other neoclouds and hyperscalers battle for grid capacity, this company is unlocking an entirely new energy source for the AI era.

It turns out that having access to a power source that no one else is tapping has advantages.

Who are they?

Crusoe.

The GPU Audio Companion Issue #14

Want the GPU breakdown without the reading? The Audio Companion does it for you—but only if you’re subscribed. Fix that here.

Company Background

Crusoe started in 2018 with a mission to eliminate energy waste.

Oil companies flare excess gas because it’s too expensive to transport. Crusoe’s solution was simple: bring the compute to the energy instead of vice versa. They built modular data centres that could be deployed at well sites, turning wasted gas into electricity and using it to power computing.

Bitcoin mining was the perfect test case.

It needed massive amounts of power, and it didn’t care where that power came from.

The solution was a success. Crusoe scaled quickly, setting up operations across the US. Flared gas was no longer a mere waste product. If you needed a cheap, reliable power source that few others had access to, it was a highly valuable asset. And then AI changed everything.

It turns out Bitcoin mining wasn’t the only energy-hungry beast out there.

Large-scale model training required far more compute and power than crypto.

AI companies were desperate for GPUs. Hyperscalers had a monopoly on supply. Traditional data centres struggled with power constraints. And Crusoe saw their next opportunity.

They started repurposing their expertise into AI cloud infrastructure, using stranded and flared energy to power high-performance compute clusters.

Now, they’re expanding their network of AI data centres, targeting enterprises and research labs that need cost-effective, sustainable computing at scale.

Executive Team

Crusoe’s leadership blends energy industry veterans with cloud and AI experts.

• Chase Lochmiller (Co-Founder, CEO, and Chairman of the Board): Leads Crusoe's strategic vision and overall operations.

• Cully Cavness (Co-Founder, President, and COO): Oversees daily operations and implements company strategies.

• Matthew DeNezza (CFO): Manages Crusoe's financial planning, reporting, and capital strategy.

• Nitin Perumbeti (CTO): Directs technology strategy, product development, and infrastructure innovation.

• Chris Dolan (Chief Data Center Officer): Oversees data center expansion and operations.

• Tara Green (Chief People Officer): Leads HR, talent development, and corporate culture initiatives.

• Tim Loos (Chief Accounting Officer): Oversees financial reporting and compliance.

• Jamey Seely (General Counse): Manages Crusoe’s legal strategy, risk management, and regulatory compliance.

The Edge

Crusoe isn’t just another AI cloud provider. Their competitive advantage is in energy efficiency and cost structure.

• Digital Flare Mitigation (DFM): Instead of burning flared gas, Crusoe converts it into electricity, lowering emissions and creating ultra-low-cost power for compute.

• Digital Renewable Optimisation (DRO): Beyond flared gas, Crusoe builds data centres near stranded renewable energy, ensuring a steady supply of cost-effective power.

• AI Cloud Compute: Bare-metal GPUs powered by direct-to-source energy pricing, offering enterprise customers cost-effective AI infrastructure.

• High-Density, AI-Optimised Data Centres: Custom-built for AI workloads, Crusoe’s facilities maximise GPU performance while reducing energy consumption. Their model is simple: Lower energy costs = lower compute costs = a long-term edge in AI cloud.

Recent Moves

• DC Provider for OpenAI’s Project Stargate: Crusoe’s Lancium Clean Campus facility in Abilene, Texas, is the first site for OpenAI’s next-generation compute project, which is set to deploy at an unprecedented scale across multiple locations in the U.S. This $3.4B AI data centre is being developed in partnership with Blue Owl Capital, and was initially leased to Oracle for AI workloads.

• $500M Fundraise: Raised a major funding round led by Founders Fund, valuing Crusoe at $3B. Supports AI cloud expansion and data centre growth.

• Vast Data Partnership: Partnered with Vast Data to deliver a high-performance, scalable data platform on Crusoe Cloud.

• SES AI, NVIDIA & Supermicro Partnership: SES AI taps Crusoe, NVIDIA, and Supermicro to accelerate battery material discovery with AI-driven molecular mapping.

Check out the Crusoe Newsroom for the latest updates.

What’s Next?

Crusoe isn’t just another AI cloud provider.

They’re rewriting the rules on how AI infrastructure gets powered.

As hyperscalers fight over limited energy capacity and regulators tighten restrictions on data centre expansion, Crusoe is tapping into an overlooked energy source. One that no one else can easily access.

Their 100MW expansion signals a major shift in the neocloud space.

Instead of relying on grid power bottlenecks, they’re unlocking a new frontier for scalable compute infrastructure.

Their partnerships with NVIDIA and Supermicro give them the hardware edge. Their Digital Flare Mitigation® technology provides a sustainability moat. And their ability to turn wasted energy into AI power makes them one of the most interesting infrastructure plays in the industry.

As demand for AI compute continues to explode, power constraints will become the defining limitation.

Crusoe’s model could be the blueprint for how AI cloud scales in a world where energy is the biggest bottleneck.

The real question?

Will hyperscalers and governments adopt Crusoe’s model? Or will they continue fighting over an increasingly constrained power grid?

One thing is clear: Crusoe isn’t waiting around for the answer.

They’re already building the future.

Keep The GPU Sharp and Independent

Good analysis isn’t free. And bad analysis? That’s usually paid for.

I want The GPU to stay sharp, independent, and free from corporate fluff. That means digging deeper, asking harder questions, and breaking down the world of GPU compute without a filter.

If you’ve found value in The GPU, consider upgrading to a paid subscription or supporting it below:

☕ Buy Me A Coffee → https://buymeacoffee.com/bbaldieri

It helps keep this newsletter unfiltered and worth reading.