Podpower Episode Atlas

Overview

In this episode of The Pomp Podcast, host Anthony Pompliano interviews Asher Genoot, CEO of HUT, a company at the forefront of the intersection between Bitcoin mining and AI High-Performance Computing (HPC). Genoot details HUT's strategic pivot from being a major Bitcoin miner to a leading provider of AI infrastructure, highlighting their unique approach of building net new projects rather than converting existing Bitcoin facilities. He explains the massive demand for AI compute, comparing it to the early days of the internet, and positions HUT as a "picks and shovels" provider for this technological revolution.

The conversation delves into the societal challenges of building large-scale data centers, addressing common fears around energy prices, water usage, and aesthetics. Genoot debunks misconceptions, explaining how data centers can actually lower energy costs and that modern cooling systems use minimal water. He emphasizes the national importance of the US leading in AI infrastructure to maintain technological leadership against countries like China. The discussion also covers innovative financing strategies, supply chain bottlenecks, and HUT's long-term vision of "physical intelligence" – redefining how technology is built in the physical world.

Genoot provides insights into HUT's business model, including their full turnkey solutions and focus on cost per token efficiency. He shares details about their large-scale, long-term contracts with investment-grade counterparties and their unique construction-phase investment-grade financing. The episode concludes with a look at the spin-out of American Bitcoin, highlighting its continued success and the importance of maintaining Bitcoin compute in North America. This episode offers a comprehensive look into the complex and rapidly evolving world of AI infrastructure and its broader implications.

Themes

AI Infrastructure Boom / The episode explores the unprecedented demand for AI compute and the critical role of data centers in supporting this growth.Societal Impact of Data Centers / Discussion on public concerns regarding data centers' impact on energy, water, and local communities, and how HUT addresses these.Strategic Business Evolution / HUT's transition from Bitcoin mining to AI HPC, emphasizing their unique development approach and long-term vision.Innovative Financing / Insights into novel financing structures for large-scale infrastructure projects, including investment-grade construction debt.National Tech Leadership / The importance of the US building robust AI infrastructure to maintain its competitive edge in the global technology landscape.

Key Concepts

01

AI HPC (High-Performance Computing)

The specialized computing infrastructure required to run complex AI models and workloads, characterized by high power consumption and specific cooling needs. HUT is building data centers designed for this purpose.

Why careThis is the 'picks and shovels' of the AI revolution, enabling the development and deployment of advanced AI technologies.

02

Energy Infrastructure Platform

HUT's mission to create a platform that supports various next-generation technologies by finding power, building infrastructure at scale, and managing energy efficiently. This includes both Bitcoin compute and AI HPC.

Why careThis diversified approach allows HUT to adapt to evolving technological demands and reduce reliance on a single technology's volatility.

03

Cost Per Token

A key metric for evaluating the efficiency of AI compute, representing the all-in cost of producing a unit of computational output. It encompasses energy, data center, and chip costs.

Why careOptimizing cost per token is crucial for competitive advantage in the AI infrastructure market, driving innovation in design and operation.

04

Investment-Grade Construction Financing

A unique financing approach where a project achieves an investment-grade credit rating during its construction phase, allowing for longer-term, cheaper capital from institutional investors, rather than short-term bridge loans.

Why careThis de-risks large infrastructure projects by securing long-term funding upfront, avoiding refinancing risks and reducing overall capital costs.

05

Physical Intelligence

HUT's long-term vision for how technology can redefine building in the physical world, using AI to design and construct infrastructure more efficiently and creatively.

Why careThis concept represents the next frontier of innovation, aiming to transform traditional industries like construction through AI-native solutions and robotics.

Quotes

"We took a very different approach. We said, we've built this platform and become one of the largest people in the world securitizing the Bitcoin network. Let's use the skill set of finding power, building infrastructure at scale, and managing power and build for the AI use case."
Asher Genoot Genoot explains HUT's strategic pivot from Bitcoin mining to AI infrastructure, leveraging existing expertise.
"The US needs to build more infrastructure if it wants to be the leader in the technology revolution around the world. And that's critical."
Asher Genoot Genoot emphasizes the national imperative for the US to invest in AI infrastructure to maintain global technological leadership.
"Energy prices actually come down with data centers coming online because you increase the utilization and you upgrade the infrastructure."
Asher Genoot Genoot debunks the common misconception that data centers necessarily raise local energy prices.
"What matters most is your cost per token, which is like your unit of compute and your cost per token on a fully depreciated and advertised basis."
Asher Genoot Genoot highlights the most critical metric for competitive advantage in the data center market.
"We were one of the first investment grade kind of construction phase data center projects in the market ever. And we were taking a bit of a risk when we did it because it was kind of a first of its kind."
Asher Genoot Genoot describes the innovative financing structure for their Riverbend project, achieving investment-grade status during construction.
"If you build an amazing underlying business, the rest will follow."
Asher Genoot Genoot shares his philosophy on long-term business success, emphasizing fundamental strength over short-term stock price fluctuations.

Chapters

010:00HUT's Pivot to AI and Infrastructure FocusAsher Genoot introduces HUT's evolution from Bitcoin mining to supporting diverse next-gen technologies, focusing on infrastructure over volatility.021:02The Insane Growth of AI DemandGenoot discusses the explosive, early-stage demand for AI, exemplified by companies like Anthropic, and HUT's role as a 'picks and shovels' provider.032:03HUT's Unique AI Infrastructure StrategyGenoot explains how HUT leverages its power and infrastructure expertise to build net new AI projects, rather than converting old Bitcoin facilities.043:04Navigating Community Concerns and US LeadershipGenoot addresses public fears about data centers (energy, water, aesthetics) and argues for the US to embrace infrastructure development for technological leadership.055:05Debunking Data Center MythsGenoot clarifies that data centers can lower energy prices and that modern cooling systems use minimal water, countering common fallacies.068:09Collocation and Energy CampusesGenoot discusses the potential for energy campuses that integrate power generation with data centers and other industrial loads for efficiency.0710:10Diverse Loads and Site ScaleGenoot explains how different compute loads (Bitcoin vs. AI) can complement each other and the future of both mega-sites and smaller, urban data centers.0813:12Full Turnkey Solutions and Cost Per TokenGenoot details HUT's approach to providing full turnkey data center solutions and the importance of optimizing 'cost per token' across the entire value chain.0917:17Securing Large-Scale AI ContractsGenoot describes HUT's process for striking multi-billion dollar, long-term contracts with investment-grade hyperscalers and AI labs.1019:19Innovative Project Financing: RiverbendGenoot explains the unique investment-grade construction financing for the Riverbend project, de-risking it for investors and securing long-term capital.1122:23Scaling Strategy and PartnershipsGenoot discusses HUT's 'pod model' for project execution and their strategic partnerships with large EPCM firms and financial institutions to support rapid growth.1224:26Development vs. Acquisition and BottlenecksGenoot outlines HUT's focus on internal development but openness to acquisitions, and identifies energy capacity, long-lead supply items, and chips as key bottlenecks.1327:29Future Vision: Physical IntelligenceGenoot shares his long-term vision for HUT in 'physical intelligence,' using AI to revolutionize how infrastructure is designed and built, and the importance of maintaining company culture during scale.1430:33Update on American BitcoinGenoot provides an update on the spun-out American Bitcoin business, highlighting its profitability despite market volatility and its role in securing the Bitcoin ecosystem.

Take-Aways

  • 01The demand for AI infrastructure is experiencing unprecedented, early-stage growth, creating a massive market opportunity.
  • 02HUT's strategy involves building net new AI infrastructure projects, leveraging their expertise in power and large-scale development.
  • 03Data centers can actually contribute to lower energy prices and utilize water efficiently with modern cooling technologies.
  • 04The US must prioritize building AI infrastructure to maintain its global technological leadership against competitors like China.
  • 05Optimizing 'cost per token' is the key competitive advantage in the AI compute market, encompassing energy, data center, and chip costs.
  • 06Innovative financing, such as investment-grade construction debt, can de-risk large infrastructure projects and secure long-term capital.
  • 07Strategic partnerships with large engineering firms and financial institutions are crucial for scaling complex infrastructure development effectively.

Open Questions

  • ?How is HUT navigating the transition from Bitcoin mining to AI High-Performance Computing (HPC)?
  • ?What are the societal concerns surrounding large-scale data center development, and how can they be addressed?
  • ?Why is it critical for the United States to lead in building AI infrastructure?
  • ?What are the most important metrics and strategies for achieving competitive advantage in the AI compute market?
  • ?How can innovative financing structures de-risk multi-billion dollar infrastructure projects?
  • ?What are the primary bottlenecks in scaling AI infrastructure, and how are companies like HUT addressing them?
  • ?What is the long-term vision for the intersection of AI and physical infrastructure, and how will it reshape industries?

Glossary

AI HPC
Artificial Intelligence High-Performance Computing, referring to the specialized computing power and infrastructure needed for advanced AI workloads.
ASIC
Application-Specific Integrated Circuit, a type of chip designed for a specific purpose, commonly used in Bitcoin mining.
Kayak
A term for the cost paid by data centers for infrastructure and system upgrades to transport energy, benefiting the overall grid.
Closed Loop System
A cooling system for data centers that recirculates a coolant without actively consuming water, similar to a home refrigerator.
LLM Models
Large Language Models, a type of AI model that can understand and generate human-like text, requiring significant computational resources.
Cost Per Token
A metric representing the all-in cost of a unit of computational output in AI, considering energy, data center, and chip expenses.
EPCM
Engineering, Procurement, and Construction Management, a project delivery method where a contractor manages all aspects of a project.
Hyperscalers
Large cloud service providers (e.g., Amazon, Microsoft, Google) that operate massive data centers and offer scalable computing resources.

People Mentioned

Asher Genoot
CEO of HUT, the guest on the podcast discussing his company's pivot to AI infrastructure.
Anthony Pompliano
Host of The Pomp Podcast, interviewing Asher Genoot.
Anthropic
An AI company mentioned for its rapid revenue growth, illustrating the demand for AI compute.
Nvidia
A US company known for its GPU chips, which are critical for AI computing.
Jacobs
A large engineering and construction management firm partnered with HUT for scaling projects.
Vera
A large company partnered with HUT for engineering, procurement, and construction management.
JP Morgan
A financial institution that led the innovative financing for HUT's Riverbend project.
Goldman Sachs
A financial institution that participated in the financing for HUT's Riverbend project.
Morgan Stanley
A financial institution that participated in the financing for HUT's Riverbend project.

Pull A Thread

  • The economics of large-scale data center development and operation.
  • The role of AI in transforming physical infrastructure and construction.
  • The geopolitical implications of AI infrastructure development and national competitiveness.
  • Innovative financing models for capital-intensive projects in the technology sector.
  • The evolving landscape of Bitcoin mining and its intersection with other high-performance computing needs.
  • Understanding the supply chain bottlenecks in the AI hardware and energy sectors.
Podpower / Atlas / 5884186