GLP Group Breaks Ground on RMB 4 Billion AI-Ready Data Centre Campus in Foshan

GLP Group has broken ground on a hyperscale data centre campus in Foshan, Guangdong Province, committing approximately RMB 4 billion in planned investment to a project designed to support up to 200 MW of IT capacity at full build-out — placing it among the largest data centre developments in China's Greater Bay Area.

The first phase, comprising 40 MW of IT capacity, is now underway. The campus is purpose-built for a single leading internet and cloud services company, expanding an existing partnership between that hyperscale customer and GLP Group. The announcement was made September 4, 2026.

Project Scale and Strategic Positioning

The Foshan campus represents a significant expansion of GLP Group's digital infrastructure footprint in southern China. With a full build-out target of 200 MW of IT capacity, the project falls within the upper tier of new AI computing campuses being developed nationally, where many of the largest facilities are targeting 120 to 200 MW or more per site.

The campus is structured as a build-to-suit, single-tenant arrangement, reflecting the economics typical of hyperscale AI infrastructure rather than speculative multi-tenant colocation.

The development adds to GLP Group's China digital infrastructure platform, which comprises approximately 20 operational and development projects with around 1.4 GW of IT capacity across key economic regions, supported by more than 700 professionals spanning site acquisition, development, design, leasing, and operations.

AI-Ready Design: Cooling, Redundancy, and Speed

The Foshan campus incorporates several technical features aimed at supporting high-density AI workloads. A hybrid air-and-liquid cooling solution is designed to support liquid-cooled deployments of up to 140 kW per rack — a specification that reflects the broader shift in China's data centre market, where newer AI training clusters are targeting 100 kW per rack or more, compared with roughly 10 kW for legacy enterprise deployments.

The facility's architecture follows an A/B dual-loop, dual-bus design with physically isolated modules on a shared base, engineered to minimize single points of failure and sustain critical workloads during maintenance or fault events. This level of redundancy aligns with what hyperscalers increasingly require for AI training environments, where idle GPU clusters carry high opportunity costs.

GLP Group is also deploying modular construction techniques — including prefabricated components and parallel on-site assembly — that the company says can reduce the construction cycle by more than 50%, enabling delivery in as fast as four months for certain phases. For hyperscale customers prioritizing speed-to-compute, accelerated delivery timelines have become a primary competitive differentiator among infrastructure developers.

The campus will also integrate GLP Group's renewable energy platform to address growing power demands, an increasingly important factor as national policy in China pushes toward low-carbon computing infrastructure.

Gavin Shen on the AI Infrastructure Opportunity

"We are delighted to expand our partnership with a leading hyperscale customer and deeply value the continued trust they have placed in GLP," said Gavin Shen, President, Digital Infrastructure, GLP China. "AI is creating one of the most significant infrastructure opportunities of the next decade and the scale of this new campus validates our strategy of building a differentiated digital infrastructure platform for the AI era. By combining infrastructure development, data centre operations, energy integration and capital management capabilities, we provide customers with end-to-end support that helps them deploy faster, scale more efficiently and create lasting value in the digital economy."

GLP Group's Broader China Digital Infrastructure Push

The Foshan development follows a series of moves by GLP Group to expand its China digital infrastructure pipeline. In July 2026, the firm signed a strategic cooperation agreement with the Ulanqab Government in Inner Mongolia to jointly develop multi-gigawatt hyperscale digital infrastructure, with approximately 1.33 million square meters of planned land area. That project involves planned investment of over RMB 30 billion across three phases, with 1.4 GW of strategic computing resources reserved.

Ulanqab has emerged as a focal point for national AI computing cluster development, with multiple gigawatt-level projects announced in the region as operators and cloud providers compete for capacity where renewable energy availability and favorable policy conditions are drawing large-scale commitments.

GLP Group's asset management arm, GLP Capital Partners, manages approximately US$79 billion of assets under management as of December 31, 2025, across logistics, digital infrastructure, and renewable energy platforms. The firm's China digital infrastructure platform also includes approximately 1.3 GW of renewable capacity under management, which it is integrating with data centre operations to address both customer sustainability requirements and regulatory alignment with China's green computing power initiatives.

The Greater Bay Area has seen surging demand for AI computing infrastructure driven by its concentration of technology companies and advanced manufacturing. The Foshan campus is designed to address that demand while extending GLP Group's presence in a market where power density, delivery speed, and energy sourcing are increasingly the primary competitive variables.

Sources

GLP Group — GLP Begins Development of AI-Ready Data Centre Campus in Foshan for Leading Hyperscaler (September 4, 2026)