San Jose, California – October 13, 2025 – MiTAC Computing Technology, a global leader in high-performance and energy-efficient server solutions, is proud to announce its participation at the 2025 OCP Global Summit (October 13–16, San Jose, CA) , booth C14. Under the theme “From AI Server to Cluster – Open for Growth. Built to Cool.”, MiTAC Computing will showcase its scalable, modular infrastructure that evolves from single AI servers to full clusters, enhanced with advanced liquid-cooling and energy-efficient designs for maximum performance and sustainability.

 

The showcase spans AI, HPC, cloud, and enterprise applications, demonstrating MiTAC Computing’s comprehensive capabilities in scaling from single servers to cluster-level integration. Teaming up with AMD, Broadcom, CoolIT, Intel, Micron, Murata, NVIDIA, and Solidigm, MiTAC Computing is accelerating the advancement of open computing and sustainable data centers.

 

From Server to Cluster | From Air-Cooled to Liquid-Cooled, From AI to HPC – Showcasing Complete Data Center Solutions

At the 2025 OCP Global Summit, MiTAC Computing is showcasing its full rack-level solutions, further reinforcing the core vision of “From AI Server to Cluster.” The company is presenting both the OCP ORv3 rack and the EIA standard rack, representing the two major directions of next-generation open architecture and traditional enterprise data centers.

 

OCP ORv3 Liquid-Cooled Rack｜Liquid Cooling and Modular Power for a Sustainable, High-Performance Data Center

The OCP ORv3 43OU liquid-cooled rack from MiTAC Computing supports up to 14 C2811Z5 multi-node servers. Each node is powered by AMD EPYC™ 9005 series processors and expandable DDR5 memory, purpose-built for high-performance computing and large-scale data center deployments. Within the overall cluster configuration, the rack integrates MiTAC Lake Erie storage modules along with management and data switches, delivering seamless collaboration across compute, storage, and networking.

 

On the infrastructure side, the rack incorporates the Murata 33kW Power Shelf MWOCES-211-P-D, providing high-efficiency and flexible power management to support sustained high-load computing environments. For thermal management, the solution features the CoolIT 200kW CHx200+ In-Rack CDU, offering 200kW-class cooling capacity to efficiently dissipate heat from high-density compute servers, ensuring stable performance and energy efficiency during continuous operation.

 

With the modular design and open architecture of OCP ORv3, MiTAC Computing’s liquid-cooled rack delivers high performance, reliability, and energy efficiency. It provides data centers with a complete upgrade path—from single server to cluster, from air cooling to liquid cooling—accelerating the adoption and expansion of AI, HPC, and cloud applications.

 

EIA Air-Cooled Rack｜Standardized Architecture with 800G High-Speed Interconnect for Rapid AI Cluster Deployment

At the 2025 OCP Global Summit, MiTAC Computing is presenting its EIA 45U air-cooled rack, configured with four MiTAC G8825Z5 8U AI servers powered by AMD Instinct™ MI350X/MI325X GPUs. This solution delivers powerful compute performance for large language model (LLM) training and generative AI inference.

 

On the networking side, the rack integrates a Dell Z9864F-ON switch featuring the Broadcom Tomahawk 5 chipset, supporting 800G high-speed interconnect to ensure low-latency and highly reliable data transmission between nodes. The rack also incorporates the MiTAC GC68C-B8056 management server and MiTAC TS70A-B8056 storage server, enabling tight integration of high-performance computing with high-speed data access.

 

By combining compute, networking, management, and storage in one unified system, MiTAC Computing extends the traditional EIA standard architecture to the cluster level. This allows enterprises to rapidly deploy compatible, scalable, and highly available AI/HPC cluster environments—without requiring changes to existing data center infrastructure—delivering on the vision of “From Server to Cluster.”

 

Live Demo｜Open Firmware Driving Transparency, Security, and Sustainability in Data Centers

At Booth C14, MiTAC Computing will also host a Live Demo showcasing how OpenBMC and Open Platform Firmware (OPF) are transforming data center management in real-world environments. In collaboration with the Open Source Firmware Foundation, ISVs, and the open hardware community, MiTAC Computing demonstrates an innovative path for replacing proprietary stacks with open firmware. Highlights include Redfish-based unified server management, optional architectures such as Coreboot, LinuxBoot, and UEFI that accelerate boot times and reduce POST by up to 50%, as well as Broadcom MegaRAID 9560-16i / 9660-16i RAID cards that further enhance data reliability and security. The solution also supports auditable security mechanisms with SBOM and NIST 800-193 protection.

 

In addition, MiTAC Computing is working with global industry leaders to advance the OCP OPF specification and is showcasing an Open Platform Firmware proof-of-concept (PoC) on AMD EPYC™ 9005/9004 processor platforms—validating its feasibility on real hardware and demonstrating its potential for future scalability. This demo offers attendees a first-hand experience of how open, modular, and sustainable infrastructure can deliver transparency, control, and long-term security for the evolution “From Server to Cluster.”

 

After showcasing its cluster-level solutions and innovations in open firmware, MiTAC Computing will also present its complete server portfolio, covering AI computing, HPC, cloud, and enterprise data processing. This comprehensive lineup highlights the company’s full capability to scale from single servers to clusters, demonstrating its strength in delivering end-to-end upgrades.

AI Computing Platforms | Liquid Cooling and GPU Acceleration for Large-Scale Training and Generative AI

• G4527G6: A 4U server based on the NVIDIA MGX™ architecture, with dual Intel® Xeon® 6767P CPUs, supporting up to 8 NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and Solidigm D7-P5520 SSDs. Combining GPU parallel computing with high-speed storage, it is designed for deep learning, computer vision, and enterprise AI development. 

• G4826Z5: Making its debut at the summit, this high-density liquid-cooled GPU platform supports up to eight AMD Instinct™ MI355X GPUs and AMD EPYC™ 9005/9004 series processors. With 24 DDR5-6400 memory slots and a maximum capacity of 6TB, the platform adopts full liquid cooling for both CPU and GPU, ensuring stable performance even under high-density deployments. Equipped with the Broadcom P2200G network adapter, it delivers high-speed, low-latency connectivity, purpose-built for large-scale AI training and data-intensive computing workloads. The G4826Z5 is featured as a flagship liquid-cooled product at this year’s exhibition.

HPC and Cloud Best Practices | OCP-Compliant Platforms for Scalable Workloads

• C2811Z5: An OCP-compliant liquid-cooled high-density multi-node server, powered by AMD EPYC™ 9005 series processors. Each node supports up to 12 DDR5-6400 memory slots with a maximum capacity of 3TB. Featuring NVMe E1.S storage interfaces paired with Micron 9550 NVMe SSDs, the system delivers stable performance even under high-bandwidth access and extended computing workloads. Designed for HPC scenarios such as scientific simulations, engineering design, and weather modeling, its liquid cooling not only enhances thermal efficiency but also helps data centers achieve optimal balance between performance, energy efficiency, and reliability.
• Capri 3: An OCP-based cloud server platform offering modularity and flexible expansion, integrated with the Broadcom N1400GD network adapter for high-speed and stable connectivity. Ideal for cloud virtualization, software-defined storage, and large-scale data lake architectures, the platform is designed to handle diverse cloud workloads and large-scale data center deployment requirements with agility.

Enterprise Data Processing & Storage | Reliable Performance and Flexible Expansion for Data-Intensive Applications

• R1520G6: An enterprise-grade 1U server powered by Intel® Xeon® 6700P series processors, combined with Micron DDR5 DRAM and Micron 6550 ION NVMe SSDs. Designed for memory-intensive applications and sustained high-density read/write workloads, the system delivers exceptional performance and durability, meeting the stringent enterprise demands for reliability and efficiency.
• R2520G6: An enterprise-grade 2U server supporting dual Intel® Xeon® 6700P series processors and up to 24 NVMe U.2 SSDs. Featuring Solidigm D7-PS1010 SSDs with PCIe 5.0 high bandwidth and long-term performance stability, it is ideally suited for data storage, big data analytics, and AI data preprocessing. The platform helps enterprises transform massive volumes of data into real-time insights, empowering smarter decision-making and operational optimization.

 

In addition to its full rack-level solutions and live demonstrations, MiTAC Computing will host two Executive Sessions on October 14 to explore the future of AI clusters and the architectural evolution of sustainable data centers.

 

MiTAC Computing sincerely invites industry leaders and partners to visit Booth C14, experience its latest server and cluster solutions firsthand, and join us in exploring the innovative journey “From Server to Cluster.”

 

For product catalogs and more information, please visit:

MiTAC Computing 2025 OCP Global Summit Landing Page

Intel Platform Brochure

AMD Platform Brochure