MiTAC showcases 96-GPU liquid-cooled rack at Shanghai AI conference

MiTAC Computing Technology, a subsidiary of the Taiwanese conglomerate MiTAC Holdings, took the stage at the 2026 World Artificial Intelligence Conference (WAIC) in Shanghai to unveil its latest server and cooling infrastructure offerings. The centerpiece of the showcase was a high-density liquid-cooled rack system that packs 96 AMD GPUs into a compact 52U form factor—a configuration designed to maximize computational density while maintaining thermal performance. Liquid cooling has become essential for modern AI infrastructure, as traditional air cooling struggles to dissipate the heat generated by dozens of GPU devices operating in parallel. The 52U footprint (equivalent to roughly 7 feet or 2.1 meters of standard server rack height) represents a significant packing density, enabling data center operators to deploy substantial AI compute capacity without requiring proportionally larger physical facilities. This demonstration occurred at WAIC, a major annual conference that brings together artificial intelligence developers, research institutions, and infrastructure vendors, making it a key venue for companies seeking to reach enterprise buyers and data center operators evaluating next-generation AI infrastructure solutions.

MiTAC Computing Technology's public demonstration of dense GPU configurations at a major AI conference underscores the infrastructure challenge facing large-scale AI deployment. As AI workloads grow in computational intensity, data centers face mounting pressure to fit more processing power into fixed physical footprints while managing heat dissipation. Liquid cooling has emerged as a key enabler for this density, allowing multiple GPUs to operate in close proximity without thermal throttling. By showcasing a 96-GPU configuration in a 52U rack at WAIC—a venue attended by AI researchers, policymakers, and infrastructure buyers—MiTAC signals both the technical maturity of its solutions and the market demand for such systems in regions like China where AI investment is concentrated.