GIBO Holdings, which describes itself as an Asia-based AI-generated animation platform, has announced a strategic initiative to develop Malaysia's first phase of a high-performance AI data centre, beginning with a 30MW deployment featuring a 14,000-GPU supercomputing cluster.
Purpose-built for industrial-scale AI development and deployment, the facility will support large language model (LLM) training, small and mid-sized model development, and large-scale inference workloads for both commercial and public-sector use cases. This development, says GIBO, represents a major step forward in strengthening Malaysia's digital infrastructure capabilities and accelerating its emergence as a major AI compute hub within the Asia-Pacific region.
The planned 14,000-GPU cluster will be designed to deliver high-density computing optimised for deep learning, multimodal AI systems, simulation workloads, and next-generation AI applications.
Once operational, GIBO says, the facility is expected to become one of the most advanced compute environments in Malaysia.
It’s also part of GIBO's long-term roadmap to scale its AI infrastructure strategy from the initial 30MW to a 100MW multi-zone AI compute campus, and subsequently to a 200MW regional flagship facility.
The broader vision includes establishing a multi-location interconnected network of AI-focused data centres spanning key regions across Malaysia. The network is intended to serve as a unified compute backbone connecting Malaysia to Singapore, Indonesia, Thailand, Japan, Korea and Greater China, creating what GIBO describes as an ‘ASEAN-to-North Asia AI compute highway’.
The data centre announced in recent days will incorporate next-generation liquid or immersion cooling methodologies designed to maximise energy efficiency and operational resilience in tropical conditions. The infrastructure will support full-spectrum model training, including trillion-parameter architectures, while reducing compute costs for enterprises and accelerating adoption of AI systems across commercial and industrial sectors.
Dedicated digital pipelines will be integrated for sector-specific applications, enabling the rapid development and deployment of specialised AI systems tailored to real-world use cases.
GIBO says it expects that the project will attract global AI companies, ecosystem partners, and next-generation technology ventures to the country.
This is the second time in recent weeks that the company’s ambitious aims for Asia have made headlines. In early December GIBO announced the signing of a core partnership agreement with Ricloud AI, a global AI cloud infrastructure provider and one of 79 official Nvidia Cloud Partners worldwide
This alliance, says the company, positions GIBO to enter the high-growth data centre and AI cloud markets, unlocking vast opportunities in Southeast Asia through strategic partnership and exclusive regional priority rights with Ricloud.
The Data Centre Dynamics website says GIBO, which launched in 2023, offers an AIGC (AI-generated content) animation streaming platform for viewers and creators. The company has said its data centres will help support the growing demand for compute power, and aims to help bring intelligent systems to various industries, in both the commercial and public sectors. GIBO also has a digital payment platform.
