At COMPUTEX 2026, Compal Electronics (Compal; TWSE: 2324) showcased its high-density AI server solutions based on NVIDIA HGX Rubin NVL8, aligning with NVIDIA’s “Seven New Chips – One AI Supercomputer” NVIDIA Vera Rubin architecture and demonstrating its engineering readiness for next-generation AI supercomputing infrastructure.

NVIDIA Vera Rubin integrates with NVIDIA Vera CPU, NVIDIA Rubin GPU, NVIDIA NVLink 6 Switch, NVIDIA BlueField-4 DPU, NVIDIA Spectrum-X Ethernet and NVIDIA ConnectX-9 SuperNIC to form a comprehensive heterogeneous computing platform. Powered by NVLink 6 switching technology, NVIDIA Vera Rubin NVL72 rack-scale configuration delivers up to 260 TB/s of total bandwidth and enables 3.6 TB/s of all-to-all bandwidth per GPU, supporting MoE models and large-scale training and inference workloads.

Compal introduces the OG231-2-L1, based on NVIDIA HGX Rubin NVL8 platform. The system features eight NVIDIA Rubin GPUs within a 2U form factor, delivering up to 400 petaFLOPS (NVFP4) to support LLM training and inference, generative AI, and HPC workloads.

Powered by NVIDIA NVLink 6, the system provides up to 28.8 TB/s of GPU-to-GPU bandwidth, enabling efficient multi-GPU scalability. The platform supports up to 2.3TB of GPU memory and 176TB/s of memory bandwidth for memory-intensive AI applications. With an optimized direct liquid cooling (DLC) design, it sustains approximately 24kW of system power, addressing the requirements of high-density AI workloads and rack-scale deployment.

Building on this liquid-cooled configuration, the portfolio is further extended with the OG430-2-L1, an ORv3 4OU liquid-cooled server platform. Designed for high power density operation, the system supports stable performance under sustained AI workloads and accommodates up to eight NVIDIA Blackwell Ultra GPUs, targeting HPC and AI workloads from training to inference.

An air-cooled (AC) configuration is also available with the SGX30-2, a 10U NVIDIA HGX B300 server system. Supporting the same GPU scale, the platform provides a deployment option for diverse data center environments while addressing AI and HPC workloads across training and inference scenarios.

These systems are deployed in a rack-level integrated configuration, combining compute nodes with a Cooling Distribution Unit (CDU) and a power rack. This configuration reflects typical high-density AI infrastructure deployments, where compute, cooling, and power subsystems are coordinated to support high power density operation.

This integrated configuration also provides visibility into key components of NVIDIA’s next-generation computing platform, allowing GPU, CPU, interconnect, and system integration to be observed within a unified environment. It offers a practical view of how next-generation AI computing architectures are implemented in real-world data center scenarios, supporting infrastructure planning and system integration.

Alan Chang, Vice President of Compal’s Infrastructure Systems Business Group, stated:“As AI workloads continue to scale, the challenge for data centers is no longer just about performance, but how to maintain stable system operation within limited space and power constraints. With the increased compute density enabled by NVIDIA HGX Rubin NVL8 platform, thermal management and deployment can no longer be effectively addressed at the individual system level. Instead, they must be considered from the perspective of the overall operating environment, where compute, power, and cooling are closely coordinated. This is also driving us to extend our design approach from individual nodes to rack-level system operation.”

As AI infrastructure continues to evolve toward higher density and larger scale, data center competitiveness is increasingly defined by system-level integration. Through its presence at COMPUTEX 2026, Compal brings together multiple system form factors along with both liquid and air cooling designs, making the rack-level operation of high-density GPU systems more visible in real-world deployment scenarios. This enables customers to progressively build infrastructure capable of supporting future AI workloads.

About Compal

Founded in 1984, Compal is a leading manufacturer in the notebook and smart device industry, creating brand value in collaboration with various sectors. Its groundbreaking product designs have received numerous international awards. 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. In recent years, Compal has actively developed emerging businesses, including cloud servers, auto electronics, and smart medical, leveraging its integrated hardware and software R&D and manufacturing capabilities to create relevant solutions. More information, please visit https://www.compal.com

Media Contact

Jack Wang Vice president and Spokesperson +886-2-8797-8588 Investor@compal.com