Seoul, South KoreaMay 6, 2026 /PRNewswire/ — MiTAC Computing Technology Corporation, a global leader in high-efficiency and energy-saving server solutions and a subsidiary of MiTAC Holdings (TWSE: 3706), today showcased its latest high-performance GPU platforms and OCP-compliant liquid cooling solutions at AI Expo Korea 2026 (Booth i21), demonstrating breakthroughs in computing density and modular efficiency.
MiTAC Computing Accelerates AI Innovation at AI Expo Korea 2026
Under the theme “Accelerate Your AI Future: Scalable Infrastructure for Next-Gen Compute,” MiTAC Computing presented a comprehensive product portfolio covering AI acceleration platforms and enterprise-grade data center solutions. Through end-to-end solutions spanning from servers to racks and racks to clusters, MiTAC Computing demonstrated its system integration capabilities for building efficient, highly scalable AI infrastructure. Welcome to the MiTAC Computing booth to explore the latest trends in next-generation computing technology.
AI Acceleration Platform | Delivering Powerful Computing Performance for AI Workloads
MiTAC Computing highlighted several GPU-optimized, highly scalable computing platforms tailored for AI training, inference, and high-performance computing applications:
G8825Z5 | High-Density GPU Server for Large-Scale AI Training
Designed for exceptional performance, it supports up to 8 AMD Instinct™ MI325X or MI350X GPUs, paired with dual AMD EPYC™ 9575F or AMD EPYC™ 9755 processors. The system delivers outstanding computing density, enhancing data processing efficiency and reducing model training cycles.
G4520G6 | GPU Platform Optimized for Generative AI and HPC
Powered by Intel® Xeon® 6 processors, it supports up to 8 NVIDIA RTX PRO™ 4500, NVIDIA RTX PRO™ 6000, or NVIDIA H200 NVL GPU Blackwell server-grade GPUs. This model offers exceptional performance per watt for cloud and HPC workloads, delivering scalable performance for generative AI, analytics, and enterprise applications.
C2811Z5 | OCP-Compliant Liquid-Cooled Multi-Node Server
Based on AMD EPYC™ 9005 series processors, the C2811Z5 is designed to maximize energy efficiency and sustainable high-performance computing. Each node supports up to 12 DDR5-6400 memory slots and offers NVMe E1.S interfaces and NVMe SSD configurations, suitable for scientific simulations, engineering design, weather modeling, and other HPC scenarios.
Capri 3 | OCP Architecture Cloud Computing Server Platform
Designed for large-scale cloud deployments, Capri 3 offers exceptional modularity and flexible scalability to meet evolving data center needs. This platform is ideal for cloud virtualization, software-defined storage, and large-scale data lake architectures.
Enterprise Solutions | Delivering Stable and Reliable Performance for Data-Driven Businesses
R2520G6 | Enterprise-Grade 2U Rackmount Server Platform
For more information, please refer to:
－ AMD Platform Brochure
－ Intel Platform Brochure
－ NVIDIA Platform Brochure