Compal presented integrated rack-level AI infrastructure solutions and applications at GTC 2026.

Compal Electronics (Stock Code: 2324) announced its participation in NVIDIA GTC 2026, taking place March 16-19 in San Jose, California at Booth 107. At the event, Compal will present its next-generation AI data center deployment architecture through a three-rack physical design. Centered on Compute, Power, and Liquid Cooling, the showcase underscores Compal's "ONE Integrated Solution" — a unified, rack-level AI infrastructure architecture. The exhibit includes high-density system designs based on NVIDIA HGX system and NVIDIA MGX architecture, featuring deployments built on NVIDIA HGX Rubin NVL8 and NVIDIA HGX B300 system.

With the introduction of NVIDIA Rubin GPUs and the rollout of NVIDIA HGX B300 and other new system architectures, compute density and power requirements continue to increase, raising the bar for data center power delivery capacity, energy efficiency, and thermal management. Against this backdrop, data center construction is progressively shifting from node-centric design toward rack-level infrastructure planning. This change in deployment scale elevates platform selection and system integration capabilities as critical factors in long-term enterprise investment decisions. The parallel evolution of NVIDIA HGX and NVIDIA MGX platforms reflects the continuity of NVIDIA's technology roadmap, and for system providers, the ability to stay aligned with this development cadence has become a key competitive differentiator.

At GTC, Compal's automotive electronics team will showcase an infrared perception system demonstrating real-time AI inference capabilities for next-generation vehicle applications. In the medical and life sciences domain, Compal will present two technical poster sessions—"GPU Annealer Molecular Docking with Fast Screening and Accurate Ranking" and "Generative Antibody Factory: GPU-Accelerated Diffusion Model & Protein Language Model for Nanobody Discovery"—focusing on accelerated molecular docking and generative antibody design. These AI workloads can be deployed on Compal's high-density GPU server platform (SX420-2A), demonstrating how rack-level compute architecture supports drug discovery and scientific AI research.

Compal will host a featured session on March 16 at 2:00 p.m., titled "Synergy at Scale: Unify Compute, Power and Cooling," where it will share its planning philosophy for high-density AI environments. Alan Chang, Vice President of Compal's Infrastructure Systems Business Group, commented, "The evolution of AI is driven not only by breakthroughs in silicon performance, but also by advancements in infrastructure-level collaboration. Through our long-standing collaboration with NVIDIA, we are positioned at the forefront of technology transitions, preparing for the next phase of deployment. As the industry advances toward higher compute density and larger-scale workloads, we aim to deliver sustained integration value across platform generations."

Founded in 1984, Compal is a leading manufacturer in the notebook and smart device industry. In 2025, Compal was recognized by CommonWealth Magazine as one of Taiwan's top 7 manufacturers and has consistently ranked among the Forbes Global 2000 companies.