The TRUTH About Bitcoin Mining & AI Data Centers
TL;DR
HUD CEO Asher Gnut explains the company's transformation from Bitcoin miner to $10B+ AI infrastructure powerhouse, detailing why ground-up energy campuses—not converted mining facilities—are critical to securing America's lead in the global AI race against China.
🔄 Strategic Pivot to AI Infrastructure 3 insights
Spun out Bitcoin mining to American Bitcoin
HUD separated its 700MW Bitcoin mining operation into a standalone public company, reducing exposure to crypto volatility while maintaining the asset as a customer.
$17 billion in net new AI contracts
Unlike peers converting old mining facilities, HUD builds ground-up AI data centers, applying their power-procurement expertise to secure massive infrastructure contracts.
Multi-technology energy platform
The company positions itself as infrastructure for 'technologies that change the world' rather than betting on a single winner, supporting both ASICs and AI workloads.
🏘️ Debunking Data Center Myths 3 insights
Energy prices decrease with proper structure
Data centers pay for grid upgrades (kayak payments) and commit to long-term consumption, which increases utilization and lowers costs for local communities.
Closed-loop cooling eliminates water usage
Modern facilities use refrigerator-style closed-loop systems rather than water-intensive cooling, consuming less water than standard office buildings.
Aesthetic design addresses NIMBY concerns
Facilities can be built quietly near homes and designed to look like museums rather than warehouses, countering community resistance that threatens US competitiveness.
⚡ Energy Campus Architecture 3 insights
Co-located generation solves grid constraints
Future campuses bring their own power generation (behind-the-meter assets) to constrained grids, allowing data centers to add capacity rather than just consume it.
Bitcoin and AI as complementary loads
Bitcoin mining (interruptible, price-sensitive) balances AI compute (bursty, less price-sensitive), allowing shared infrastructure with flexible power consumption.
Distributed deployment strategy
While building gigawatt-scale campuses, HUD also deploys smaller 50-100MW sites closer to urban areas as power constraints make mega-sites harder to permit.
🔗 Full-Stack Integration 3 insights
Optimizing cost per token
Competitive advantage comes from minimizing fully-loaded compute costs across the entire stack: energy procurement, facility construction, and chip orchestration.
Turnkey solutions beyond powered shells
HUD builds complete data centers including IT infrastructure and owns four power plants, controlling the value chain from generation to compute.
Vertical integration via High-rise.ai
The company incubated a neocloud business to understand GPU utilization patterns, ensuring infrastructure design aligns with actual AI workload requirements.
Bottom Line
Winning the AI infrastructure race requires building ground-up energy campuses that co-locate power generation with compute, while addressing community concerns through transparent education about grid benefits and aesthetic design—treating data center construction as national security infrastructure rather than simple real estate development.
More from The Pomp Podcast
View all
Why Are Bitcoin & AI Stocks CRASHING?!
Despite recent volatility in AI stocks and Bitcoin, macro hedge fund manager Jordy Visser argues we are only in the second inning of AI's infrastructure buildout, with memory shortages and physical bottlenecks creating strategic investment opportunities rather than signaling a bubble burst.
Will The K-Shaped Economy Destroy America?
Darius Dale argues the Fed is conducting 'financial repression' by maintaining hawkish rhetoric while effectively abandoning the 2% inflation target to avoid market panic. He identifies monetary drivers like deficit spending and credit growth as true inflation predictors rather than expectations, while warning that a K-shaped economy is pushing consumer delinquencies toward crisis levels despite aggregate resilience.
I Just Revealed My Current Portfolio…
Anthony Pompliano argues that MAG 7 stocks have sold off due to overblown fears about inflation and AI infrastructure spending, creating an attractive entry point for long-term investors. He believes inflation will ease in late 2025, AI demand remains robust despite efficiency improvements, and quality businesses should be bought when they temporarily fall out of favor.
The Biggest Pivot In AI History Is Happening Right Now
Jordi Visser argues that recursive self-improvement in AI has arrived years ahead of schedule, triggered by a jailbroken Anthropic model and immediate government intervention, while Chinese open-source alternatives and algorithmic efficiency gains threaten to invalidate the hyperscalers' massive data center capex investments.