China Mobile’s $14B Compute Bet: 7 Gigawatt AI Data Centers by 2027
TREE NEWS reports: China Mobile plans to invest more than 100 billion yuan (about $14 billion) in computing power by 2027, chairman Chen Zhongyue said at the company’s 2026 mobile network conference on November 6. The telecom giant is building three new gigawatt-scale computing parks this year, expects to deliver 1.5 gigawatts of capacity, and will add two more next year to reach seven gigawatt-class data parks in total.
Chen said China Mobile has already built multiple 10,000-card intelligent computing clusters and that its compute capacity will double for two consecutive years. Total procurement is set to exceed 500 billion yuan by 2027, with compute accounting for over 100 billion yuan. The company also operates six data centers overseas.
Why It Matters for Crypto and AI
The announcement underscores how state-backed telecoms are becoming primary infrastructure providers for AI — a trend with direct implications for decentralized compute networks. As hyperscale capacity concentrates in the hands of a few operators, crypto-native GPU marketplaces and decentralized training networks will increasingly compete on cost, censorship resistance, and access for smaller developers.
- Compute as an asset class: Gigawatt-scale buildouts reinforce the idea that compute is scarce and monetizable, supporting tokenization models for GPU capacity and inference.
- Supply chain pressure: Massive procurement plans could tighten global supplies of advanced chips, GPUs, and networking gear, affecting decentralized networks that source hardware on open markets.
- Geographic arbitrage: China’s buildout may push decentralized compute providers to focus on regions where state capacity is thinner, or to partner with telecoms for last-mile distribution.
The Bigger Picture
China Mobile’s pivot from connectivity to compute mirrors a global shift: telcos are repositioning as AI infrastructure players. For the crypto sector, the key question is whether decentralized compute can scale fast enough to remain a credible alternative to centralized hyperscalers. The next two years — exactly China Mobile’s doubling window — will be decisive.
