Envision Greenwise’s Latest Pivot To AI Infrastructure After Construction, Recycling

After reinventing itself from a construction contractor to an environmental services and recycling platform, Envision Greenwise is making another bold turn—this time into AI infrastructure. The company is pouring capital into high-performance servers and data center capacity as it chases faster growth and higher margins than those available in traditional recycling and engineering services.

Key Points

• Nearly 1.7 billion yuan committed to server purchases to accelerate AI computing capacity.
• Targeting 16 MW of near-term computing power and 40% of revenue from intelligent computing by March next year.
• Financing the push with new share placement and convertible bonds totaling roughly HK$1.1 billion.

From Construction Roots to Recycling

Envision Greenwise’s story stretches back to 1985, when Hong Kong entrepreneurs Chan Kam Tong and his cousin Chan Kam Ming acquired Head Fame, a building contractor later renamed Golden Ponder. After decades focused on superstructure projects, the company went public in 2018. But the sector’s chronically low valuations, coupled with pandemic-related disruptions and a weak property market, weighed heavily on the business.

In 2022, industry veteran Kwok Chun Sing stepped in, ultimately taking control through a reverse takeover. With more than two decades in environmental technology and new energy, he set out to reposition the company as a green-energy and circular-economy platform. Anticipating rising demand from new energy vehicles, Envision Greenwise expanded into battery recycling, energy storage systems, and EV charging.

The strategy delivered rapid top-line growth: revenue climbed to HK$2.46 billion in the latest fiscal year, with profit of HK$61.03 million. However, margins remained thin—gross margin fell to 8%—highlighting the limits of profitability in many environmental and recycling projects.

The Big Bet on AI Infrastructure

Seeking stronger returns, Envision Greenwise is now steering into AI infrastructure, a segment benefiting from surging demand for training and inference of large models. A planned acquisition of Shanghai Yovole Cloud Calculation marked a decisive step, adding cloud computing, data center operations, and intelligent computing services to the portfolio. Critically, Yovole’s licenses and procurement qualifications help secure high-performance GPUs, the scarce backbone of modern AI workloads.

At present, the company can supply around 1.5 MW of computing power to clients. Management’s near-term target is 16 MW, with a goal of making intelligent computing contribute 40% of total revenue by March next year. To move quickly, Envision Greenwise has been on a buying spree: it announced an investment of 380 million yuan for over 100 servers, then followed with a 1.29 billion yuan purchase of roughly 400 more. In aggregate, it has committed nearly 1.7 billion yuan to server capacity.

Funding the Shift

Scaling AI infrastructure is capital intensive. As of the end of March, Envision Greenwise held about HK$230 million in cash. To fund the AI buildout, the company placed 118 million new shares at HK$4.66 per share, raising approximately HK$550 million, and issued nearly $70 million in convertible bonds. Combined, the financing round brought in close to HK$1.1 billion.

These moves underscore both management’s urgency and the financial demands of competing in AI infrastructure, where capacity must be secured before clients can be fully onboarded and monetized.

Opportunities and Risks

The strategic logic is clear: AI infrastructure sits in a fast-expanding market with pricing power tied to scarce GPUs and robust demand from model developers and enterprise adopters. If Envision Greenwise can scale quickly, secure reliable power and cooling, and maintain access to cutting-edge chips, it could unlock far higher margins than in recycling and construction.

But the risks are equally notable. The company is a newcomer to AI computing and must prove it can operate data centers, manage GPU fleets, and deliver service-level agreements at scale. Frequent strategic shifts—from construction to recycling to AI—raise questions about execution focus. Capital intensity is another hurdle: ramping from 1.5 MW to 16 MW requires sustained investment, even as the industry’s technology cycles and customer requirements evolve rapidly.

Partnerships will likely matter. Supplier relationships, cloud service licenses, and channel access can accelerate customer acquisition and hardware procurement, but they also require careful management and ongoing investment.

Market Reaction and Outlook

Investor sentiment has been volatile. Shares rallied in May as the AI narrative gained traction, climbing from around HK$3 to HK$6.42 within a month. After the acquisition announcement, the stock retreated and has since traded with wide swings, recently hovering just above HK$4. The whipsaw reflects both enthusiasm for the AI theme and caution toward companies in transition.

Looking ahead, milestones to watch include the pace of server deployments, the ramp to 16 MW, customer wins that translate to recurring revenue, and evidence that intelligent computing can reach the targeted 40% of total revenue in the stated timeframe. Clear disclosure on utilization rates, power availability, and chip procurement would further clarify the strength and durability of the pivot.

Envision Greenwise has shown an ability to raise capital and reinvent itself. Whether this latest transformation can deliver sustained, profitable growth will depend on disciplined execution in a fiercely competitive, capital-hungry arena.