Cango Inc., the NYSE-listed Bitcoin mining company operating under the ticker CANG, has announced that its high-performance computing and AI inference subsidiary, EcoHash Technology LLC, has reached two significant operational milestones: completing infrastructure modifications at its owned 50-megawatt facility in Georgia and beginning commercial GPU compute services using an initial batch of GPU servers now in operation.

Infrastructure Overhaul at the Georgia Site

EcoHash has completed the infrastructure modifications and retrofits needed to establish a dedicated AI compute section within Cango's existing Georgia facility.

The section is designed to support up to three megawatts of capacity, with room identified for future expansion.

The build-out involved the delivery, installation, and testing of high-density compute containers, which have begun their operational rollout through the power-on and activation of an initial batch of GPU servers.

The containerized data center solution is described by the company as a modular building block for AI compute, standardizing power, cooling, networking, and operations into a repeatable unit. Each module is engineered to handle sustained high-density GPU loads.

Power is delivered through an industrial-grade transformer setup, a precision cooling system supports the site's compute density, and network connectivity is provided through fiber infrastructure designed specifically to handle AI compute workloads.

Because the compute modules were deployed alongside Cango's existing Bitcoin mining infrastructure at the same Georgia location, EcoHash said it was able to minimize new construction and site retrofit work, accelerating the timeline from module delivery to power-on.

The modular approach is intended to allow the company to scale AI infrastructure faster and more cost-effectively across suitable sites in the future.

Commercial Services Now Underway

On the commercial side, EcoHash has begun providing GPU compute services to paying customers using part of its initial batch of GPU servers.

The company noted that additional GPU deployment remains ongoing. The announcement describes this as the initial commercial utilization of EcoHash's GPU capacity at the Georgia site.

The initial customer and prospective customers identified by EcoHash include GPU cloud platforms and AI-native cloud infrastructure providers.

The company also disclosed that it is in discussions with additional prospective clients across other categories of AI compute users, including decentralized compute networks and advanced AI research laboratories.

Simon Tang, Chief Financial Officer of Cango Inc., said in a statement that powering on the first batch of GPU servers and beginning commercial GPU compute services represents an important step for EcoHash.

Tang added that by utilizing the existing infrastructure and energy access at the Georgia facility, the company is demonstrating what it describes as a plug-and-play container solution in a commercial setting.

He described the operational rollout as a strategic proof-of-concept hub as the company works toward replicating the model at future sites.

Background on EcoHash and Its Role Within Cango

EcoHash Technology LLC was founded in 2025 as the dedicated high-performance computing and AI inference arm of Cango Inc.

The subsidiary deploys standardized, plug-and-play compute modules with the goal of delivering scalable AI inference solutions.

The company says it is developing a lightweight orchestration layer intended to function as the network's central nervous system, with the objective of dynamically provisioning compute power to meet real-time workload demands. Cango Inc. itself describes its broader ambition as building an integrated, global infrastructure platform capable of powering what it calls the future digital economy.

The company entered the digital asset space in November 2024 and has since activated pilot projects in both integrated energy solutions and distributed AI computing.

Its Bitcoin mining operations span North America, the Middle East, South America, and East Africa. Cango also continues to operate an online international used car export business through its AutoCango.com platform.

Strategic Significance of the Georgia Deployment

The Georgia facility serves as the first operational demonstration of EcoHash's containerized AI compute approach.

By co-locating the AI compute modules with existing Bitcoin mining infrastructure, Cango avoided the time and cost associated with building a greenfield data center.

The company has positioned the Georgia site as a template it intends to replicate at other locations as it pursues growth in the AI compute market.

The three-megawatt AI compute section represents a portion of the facility's total 50-megawatt capacity, which continues to support Cango's Bitcoin mining operations alongside the new AI workloads.

The dual-use nature of the site reflects Cango's stated strategy of leveraging its existing energy access and operational infrastructure to expand into high-performance computing without requiring entirely separate facilities.

The company has not disclosed financial terms related to its initial GPU compute customer arrangements, nor has it provided specific revenue projections tied to the Georgia deployment or future sites under consideration.