Cango said its subsidiary EcoHash Technology LLC has begun commercial operations and launched an official digital portal at www.ecohash.com as the main access point for its high-performance computing and AI inference business. The company positioned the platform around two target groups: AI developers looking for low-latency compute close to power sources, and compute operators with heavy energy demands seeking modular options to diversify infrastructure.
Power demand gap frames the market pitch
Citing Goldman Sachs Research, Cango said U.S. data center electricity demand could reach 700 TWh by 2030, driven largely by AI inference workloads, while maximum available supply remains only a little above 300 TWh. That implies a structural gap of roughly 400 TWh. EcoHash says it is addressing that mismatch by using Cango’s global energy footprint to deploy standardized plug-and-play compute modules together with its proprietary EcoLink Orchestration Platform.
The orchestration system is designed to unify and schedule geographically distributed compute capacity and support enterprise-grade uptime through intelligent failover. In Cango’s description, the result is elastic, low-latency compute that can scale and be activated on demand. The emphasis is on linking scattered energy resources with distributed compute capacity rather than relying on a single site.
Georgia 50MW mining facility to serve as live showcase
Cango is allocating space at its owned 50MW mining facility in Georgia for the initiative. Using the site’s existing infrastructure and energy access, the facility will host full-series container models as a “living showroom.” The company said the site is intended to demonstrate real-world performance under different thermal and power configurations while also acting as a proof-of-concept hub for collaborators across digital infrastructure and the mining sector.
By presenting the commercial use case for these plug-and-play modules, Cango said it wants to bring global partners into the EcoHash network and replicate the Georgia model across other high-potential locations, both inside and outside its current footprint.
EcoHash CTO says commercialization is accelerating
EcoHash Chief Technology Officer Jack Jin said the subsidiary is the core vehicle for Cango’s future-ready platform strategy and its next growth engine, and that it is now entering a phase of accelerated commercialization. He said the company’s orchestration layer functions as the central system of the network, enabling intelligent real-time resource allocation that connects decentralized energy assets directly with demand from LLM inference, generative AI, and other compute-intensive applications as node infrastructure scales.