Podcast Summary: "Bitcoin Miners Are Pivoting to AI. How Does It Impact Crypto?"
Unchained — Bits + Bips, Ep. 943 (Nov 8, 2025)
Host: Laura Shin (for Bits + Bips: Steve Ehrlich)
Guests: John Todaro (Managing Director, Needham & Co.) and Kevin Dede (Senior Research Analyst, H.C. Wainwright)
Main Theme
This episode dives into the evolving relationship between Bitcoin mining and the explosive rise of AI and high-performance computing (HPC) infrastructures. As traditional mining profitability declines, miners are seeking new revenue streams—primarily by repurposing their infrastructure for AI computing power. The panel explores the drivers, risks, and the future implications for both the crypto and data center sectors as these industries become more intertwined.
Key Discussion Points & Insights
1. The Shift from Bitcoin Mining to AI & HPC (04:22–14:00)
- Industry-wide Pivot: Both John and Kevin note that nearly every significant bitcoin miner is now actively exploring or executing a pivot to AI and HPC, driven by record demand for computing capacity.
- Notable Quotation:
"Almost all the investors we're speaking with, it's around HPC, AI, workload opportunity around the miners." — John (04:22)
- Catalysts for the Shift: Major IPOs like CoreWeave, large-scale contracts (notably Oracle's massive RPOs), and deals with cloud hyperscalers (OpenAI, Google, Microsoft) have "rerated" miners’ valuations.
- Defining HPC & Infrastructure Terms (05:24–07:31):
- Gigawatt: "A massive amount of power… the US data center market today is like 25 gigawatts… when you’re talking 1, 2, 3 gigawatts it’s significant." — John (06:04)
- Neoclouds: GPU-as-a-service providers (CoreWeave, Lambda Labs, Nebius) that lease GPU capacity to end customers.
- RPO (Remaining Performance Obligations): Measures customer backlog for data center capacity.
2. Stock Performance & Miner Strategies (09:08–14:43)
- Market Divergence: Mining company indices are outperforming Bitcoin itself, up 120% over three months despite Bitcoin’s price slide—driven by "AI Hopium" and data center demand.
- Not All Miners Equal: Only larger, capitalized miners can pivot to AI/HPC; smaller miners are being left behind.
- Splitting Operations & Evolving Business Models: Companies are spinning off HPC divisions; for example, Hive (BuzzBit Digital, WhiteFiber) and Core Scientific are transitioning existing facilities.
3. Risks, Bubbles, and End Game Uncertainties (13:52–16:31)
- Execution Risk: The biggest question is whether this boom is sustainable, or a bubble constrained by how much AI compute demand will ultimately exist.
- “How long does it last? Is it a bubble? What are the implications if the market does become satiated and how… does the end market actually pay for the huge amount of investment?” — Kevin (12:44)
- Historical Perspective: Parallels are drawn to the late-1990s tech boom and the unpredictable seismic changes brought by new computing paradigms.
4. Hype vs Substance: Investor Cautions (16:31–20:55)
- "Long Island Iced Tea" Companies: Not all miners overpromising pivots to AI can deliver—investors should scrutinize actual power pipelines, execution capability, and the nature of customer contracts.
- “Not all megawatts are created equal… due diligence the pipeline.” — John (17:44)
- Business Model Comparisons:
- Colocation Model: Longer-term, more stable, less risky (e.g., 15-year contracts with quality clients).
- GPU-as-a-Service: Shorter-term, more competitive and carries higher risk due to rapid hardware obsolescence and aggressive debt-fueled expansion.
5. Economics of Mining and Data Centers (29:41–44:00)
- Declining Profitability: Hash price index at all-time lows; mining remains only viable due to more efficient machines and cheap power pockets.
- “Each bear market brings you a lower hash price. That’s because machines are more efficient.” — Kevin (31:47)
- Geographic & Business Flexibility:
- Bitcoin mining can operate anywhere cheap power and a data/internet link exist ("all you need is a Starlink terminal").
- Data center/HPC deployments require fiber, water, and infrastructure—more limited to developed regions.
- Behind-the-Meter Advantage (36:40–38:56):
- Miners sourcing power directly from generation sites avoid grid costs ("behind the meter"), which can give access to sub-$0.02/kWh rates.
6. Capitalization and Financial Engineering (48:20–51:10)
- New Funding Models:
- Most new convertible bonds, project financing, and ABS (asset-backed securities) are now used to fund HPC expansions (not mining).
- Customer contracts can sometimes serve as collateral.
- “The debt market exists in support of these companies purely on their HPC opportunity.” — Kevin (50:21)
7. Policy and the Future US Role (51:10–55:11)
- US Mining Outlook:
- While AI/HPC is absorbing much of the highest value data center space, bitcoin mining isn’t vanishing—cheap and remote power in the US will likely always support some level of mining.
- Political interest, e.g., Trump family involvement, could drive more efforts to keep mining “onshore.”
- Notable Quotation:
“There’s probably always going to be a place for bitcoin mining. I was just making the case the infrastructure in the US is very much set up more for AI than bitcoin mining.” — John (53:51)
8. Possible Coexistence and Future Proof-of-Work Assets (55:13–59:06)
- Coexistence:
- Concept of "mullet miners": facilities running both HPC (front) and bitcoin mining (back), depending on load and market factors.
- Potential for dynamic sharing of capacity, though technical and economic challenges remain.
- Mining Other Coins:
- Most industrial miners only consider altcoin mining when highly profitable; strong Bitcoin-Maximalist ethos persists.
9. Government Involvement and National Security (59:06–61:51)
- Potential Federal Involvement:
- Speculation about whether US government could take stakes in major miners to ensure national network participation and security.
- Decentralization is still crucial; risks if too much hashpower leaves the US.
Memorable Quotes & Moments
- “I don’t think we have a clue about what artificial intelligence can bring to us. If you look out the next 30 years, life’s going to be not incrementally different, but diametrically different.” — Kevin (02:02, repeated at 15:03, 14:58)
- “Not all megawatts are created equal.” — John (17:43)
- “Right now it’s the Oklahoma land rush. Anybody who can scrap together 100 megawatts can make some money, sign up a deal. But what happens longer term?” — Kevin (22:38)
- "All you need for a bitcoin mine is electricity and a Starlink terminal from Elon, and you’re good to go." — Kevin (35:24)
- "I think there's probably always going to be a place for bitcoin mining." — John (53:51)
Timestamps for Key Segments
| Timestamp | Topic | |-----------|-------| | 04:22 | Rise of AI/HPC demand for bitcoin miners | | 09:08 | Mining stocks vs. Bitcoin price divergence | | 13:36 | Risks, bubbles, and long-term sustainability | | 16:31 | Hype signals & what investors should look for | | 29:41 | Economics: hash price, hash rate, mining costs | | 36:40 | Power sourcing: behind-the-meter explained | | 48:20 | Debt/converts for funding HPC expansions | | 51:10 | US hash rate, policy implications & outlook | | 55:13 | "Mullet miners" and hybrid HPC/BTC sites | | 57:26 | Mining altcoins—future prospects | | 59:06 | Speculating on federal government involvement | | 62:46 | Closing remarks and contact info |
Concluding Thoughts
The episode captures a pivotal moment: bitcoin miners, rather than fading with declining profitability, are repurposing their unique infrastructure for the AI and HPC gold rush. The panel cautions against hype, advocates deep diligence, and forecasts a world where mining, cloud, and AI boundaries blur. The US may cede some hash rate share to HPC, but mining will persist where it makes sense. Meanwhile, public policy, geopolitics, and market demand will shape how these industries evolve together.
Contact Info
- Kevin Dede: KD[at]hcw.com
- John Todaro: jtodaro[at]needhamco.com
For anyone interested in the future of crypto mining or the data center economy, this episode delivers valuable, expert-driven context and forward-looking analysis.
