1. Reconstruction of underlying logic: Strategic reassessment of the value of electric power resources
Bitcoin mining companies are undergoing a paradigm revolution from 'hash power producers' to 'AI infrastructure providers':
Scarcity of power generation capacity: American mining companies control 6.3GW of operational capacity (accounting for 38% of global Bitcoin mining power), and their layout of low-cost, high-stability energy hubs (such as Texas) has become an ideal location for AI training centers.
Reusability of infrastructure: The power supply systems of mining farms (such as substations and cooling facilities) can be directly transformed into AI server clusters, with transformation costs being only 1/3 of building new data centers.
Policy arbitrage opportunities: The U.S. (Inflation Reduction Act) provides tax credits for clean power (up to 30%), combined with Bitcoin mining being categorized as 'critical infrastructure' (recognized by the Department of Homeland Security in 2025), forming a closed-loop policy dividend.
Two, paradigm shift in market pricing
1. AI-driven reconstruction of valuation models
Traditional model: EV/EBITDA (Enterprise Value/Earnings Before Interest, Taxes, Depreciation, and Amortization)
New model: EV/TPU-H (Enterprise Value/per second trillion operations)
Case: IREN Ltd, in collaboration with NVIDIA, transformed its mining site into a DGX SuperPOD, with its valuation reconstructed from 0.8x EV/EBITDA to 12x EV/TPU-H.
2. Structural migration of capital flow
Crypto-native funds: Grayscale Bitcoin mining ETF (GBTC) experienced a net outflow of $1.5 billion, but the scale of crypto hedge funds shifting to mining stocks (e.g., CIFR, IREN) reached $2.3 billion.
Traditional institutional funds: Asset management giants like BlackRock and Invesco will include mining companies in 'AI infrastructure' thematic funds, with targets including CIFR, IREN, and Core Scientific, which has not yet gone public.
3. Value transmission of power-computing-algorithms
Electricity → computing power → algorithms → data products
Mining companies sell electricity ($0.05/kWh) → exchange for AI training computing power ($0.2/kWh) → participate in AI model revenue sharing (e.g., IREN's profit-sharing agreement with Anthropic).
Three, innovative paths of technological integration
1. Hybrid architecture of mining machines and servers
Bitmain's Antminer S19j Pro mining machine can run SHA-256 hash calculations and AI inference tasks simultaneously by modifying the cooling module, achieving a 40% increase in computing power reuse rate.
2. On-chain energy trading
Mining companies issue ERC-3643 energy tokens (e.g., CIFR's CF-Power Token), fragmenting electricity production capacity and allowing AI companies to collateralize computing power for liquidity through DeFi protocols (e.g., Aave V3).
3. Distributed deployment of edge AI
Deploy NVIDIA Jetson edge servers at mining sites, utilizing the redundant bandwidth of the mining site (average 1.2Gbps) to provide real-time inference services, forming a 'power-computing data' trinity edge node.
Four, risks and challenges
1. Transmission risk of policy fluctuations
The differences in energy policies across various states in the U.S. (e.g., New York state prohibiting new mines) may hinder capacity expansion.
The ESG controversy of Bitcoin mining (carbon footprint issues) may affect institutional allocation willingness.
2. Substitution risk of technological paths
Quantum computing breakthroughs may render the SHA-256 algorithm ineffective, forcing mining companies to shift to Post-Quantum encryption (e.g., Sidh protocol), incurring technology migration costs.
3. Involution of market competition
Traditional data center operators (e.g., Equinix) are beginning to lay out AI-specific power facilities, directly competing with mining companies.
Five, investment strategy recommendations
1. Selection criteria for targets
Capacity quality: Preferentially select mining companies with low electricity prices (<$0.06/kWh) and a high proportion of renewable energy (>70%).
Technical collaboration: Focus on companies that sign exclusive agreements with GPU manufacturers like NVIDIA and AMD (e.g., IREN).
Tokenization degree: Assess the liquidity of its energy tokens in the DeFi ecosystem (e.g., CIFR's CF-Power Token trading volume on Uniswap V4).
2. Innovation in trading tools
Directly trade U.S. stocks/Hong Kong stocks using USDT on BiyaPay, utilizing its zero Maker fee advantage for high-frequency arbitrage.
Hedge electricity price volatility risk through CME Bitcoin mining futures contracts (newly launched in 2025).
3. Cross-market arbitrage opportunities
Capture the divergence in correlation between mining company stocks and Bitcoin prices: When CIFR's stock price rises more than 30% above BTC, short CIFR and go long BTC.
Conclusion: The Web3 narrative of the energy revolution
The reconstruction of Bitcoin mining companies' valuations is essentially a Web3 narrative of the integration of energy revolution and digital economy. As electricity shifts from 'mining cost' to 'computing asset', these companies are writing a new value equation:
Mining site = energy bank + computing cloud + edge data center
In the future, its valuation may no longer be anchored to Bitcoin prices but deeply bound to new variables such as AI training costs, edge computing demand, and renewable energy premiums. This revolution is not only about investment returns but also signifies a paradigm shift in energy infrastructure—electricity is becoming the new oil of the digital age.
