EcoHash Expertise LLC, the devoted HPC and AI inference subsidiary of Cango Inc. (NYSE: CANG), launched its public digital portal on 13 April 2026, asserting the beginning of business operations.
It additionally unveiled plans to function a portion of its 50-megawatt (MW) Georgia mining facility as a dwell proof-of-concept hub for the AI compute business.
What’s EcoHash, and why is it coming into the market now?
Cango (CANG) based EcoHash in 2025 as a part of its aim to transform the corporate’s world power infrastructure right into a distributed AI compute community. EcoHash’s industrial launch targets AI builders searching for low-latency, near-source compute capability, and energy-intensive compute operators searching for modular pathways to infrastructure diversification. Cango (CANG) believes the latter is underserved by standard information middle suppliers.
This growth is coming at a time when researchers from Goldman Sachs are forecasting that U.S. information middle energy demand may attain 700 TWh by 2030, and this can be pushed predominantly by AI inference workloads.
Nonetheless, the present out there provide stays simply above 300 TWh, leaving a spot of about 400 TWh whilst compute demand steadily will increase.
That is the industrial rationale EcoHash is constructed round, and it was identified by Cango’s CEO Paul Yu, who calls the “Energy Hole” the disconnect between rising AI compute demand and constrained grid capability.
In line with Jack Jin, chief expertise officer of EcoHash, “EcoHash represents the core car of our technique to architect a future-ready platform and function our subsequent progress engine, now coming into a section of accelerated commercialization.”
The subsidiary’s industrial launch follows a interval of intensive capital deployment. In April 2026, Cango (CANG) introduced the completion of two financing transactions totaling $75 million, a $65 million fairness shut from board insiders Xin Jin and Chang-Wei Chiu, and a $10 million convertible observe from Hong Kong-listed DL Holdings Group Restricted (HKEX: 1709).
Cango (CANG) additionally entered a memorandum of understanding with DL Holdings for as much as $10 million in additional co-investment.
These transactions adopted an earlier $305 million increase from the sale of Bitcoin holdings used to retire debt and reset the stability sheet.
What’s the Georgia facility designed to display?
EcoHash’s launch technique is backed by the Cango-owned 50MW Georgia mining facility, the place the corporate is dedicating house to function full-series container fashions as what it describes as a “dwelling showroom”.
The location is engineered to display real-world efficiency throughout various thermal and energy configurations, functioning as a strategic proof-of-concept hub for business collaborators throughout the digital infrastructure and mining ecosystem.
Cango (CANG) intends for a portion of the Georgia facility to function the replicable template for a globally distributed AI compute community, with ambitions to scale the mannequin throughout high-potential websites each inside and past its present mining areas spanning North America, the Center East, South America, and East Africa.
The industrial viability of its plug-and-play modules in Georgia will allow the corporate to draw world companions into the EcoHash community, operators who can combine present infrastructure into the platform as an alternative of constructing new information facilities from scratch.
How does the EcoLink platform come into the image?
The operational spine of EcoHash is the proprietary EcoLink Orchestration Platform, a software program layer that unifies and schedules geographically dispersed compute capability throughout the community.
EcoLink is constructed to ship enterprise-grade uptime by means of clever failover, provisioning compute energy to fulfill real-time workload calls for.
It’s the mechanism that transforms a set of repurposed mining websites into one thing resembling a traditional hyperscale providing.
In his remark, Jin said that EcoLink is “the central nervous system of our community”, constructed to allow clever, real-time useful resource allocation connecting decentralized power property on to the calls for of enormous language mannequin inference, generative AI, and a rising vary of compute-intensive purposes.
The consequence, per Cango (CANG), is elastic, low-latency compute that scales on demand, with out the capital expenditure and multi-year lead occasions related to constructing new information middle capability.
