MiTAC Computing Advances Agentic AI Infrastructure with 6th Gen AMD EPYC™ Server CPUs

SAN FRANCISCO, July 23, 2026 /CNW/ -- MiTAC Computing Technology Corp., a global leader in high-performance, energy-efficient server solutions and a subsidiary of MiTAC Holdings Corporation (TWSE:3706), is highlighting its latest agentic AI infrastructure at AMD Advancing AI . MiTAC's showcase focuses on its premier exhibition lineup, featuring its latest liquid-cooled and air-cooled AI servers and next-generation multi-node platforms built on the new 6^th Gen AMD EPYC Server CPUs. As part of the launch, MiTAC is introducing a portfolio of next-generation platforms designed to help customers scale AI, cloud, HPC and enterprise workloads with greater performance, efficiency and flexibility.

MiTAC Computing Advances Agentic AI Infrastructure with 6th Gen AMD EPYC Server CPUs

The new MiTAC M2810Z6 and M2610Z6 systems, powered by  6^th Gen EPYC Server CPUs, reflect MiTAC's commitment to delivering infrastructure choices that align compute density, memory bandwidth, and I/O capability with evolving data center requirements. With support for AMD's Venice "Zen 6" architecture, PCIe Gen 6 I/O subsystem, and CXL 3.1, these platforms are engineered to meet the demands of modern AI-driven environments while optimizing power and system density.

"Agentic AI and next-generation enterprise workloads require infrastructure that is fast, scalable and efficient," said Rick Hwang, President of MiTAC Computing Technology. "By launching our latest platform powered by 6^th Gen AMD EPYC Server CPUs, MiTAC is offering customers the building blocks to deploy AI-ready infrastructure while supporting the workloads that run their business today. "

Product Highlights

MiTAC M2810Z6 – 2U 2-node, single-socket 6^th Gen AMD EPYC SP7 multi-node server
The M2810Z6 is a high-density 2U 2-node platform built around 1-socket AMD SP7 architecture per node. Each node supports a 6th Gen AMD EPYC Server CPU up to 600W cTDP, 16 DDR5 DIMM slots with up to 4TB of DDR5-8000 memory, and flexible front hot-swap NVMe storage options including E3.S, E1.S, and U.2. The platform also includes OCP 3.0 TSFF expansion to support AI, HPC, and cloud-scale deployment needs.

MiTAC M2610Z6 – 2U 2-node, single-socket 6^th Gen AMD EPYC SP8 multi-node server
The M2610Z6 is designed for enterprises seeking a cost-efficient and versatile platform for virtualization, storage, edge, and telco workloads. Featuring 1-socket AMD SP8 architecture per node, the system supports AMD EPYC™ 9006 SP8 server processors up to 400W cTDP, 16 DDR5 DIMM slots, up to 4TB of DDR5-6000 memory, and OCP 3.0 TSFF connectivity. Like the M2810Z6, it also offers flexible hot-swap NVMe storage options to support a wide range of deployment scenarios.

"Enterprise AI has moved from experimentation to production, and agentic AI will place even greater demands on compute, memory bandwidth, efficiency and system-level orchestration," said Dan McNamara, senior vice president and general manager, Compute and Enterprise AI, AMD. "With 6^th Gen AMD EPYC Server CPUs, AMD is helping customers scale enterprise AI with the performance and efficiency needed to support increasingly complex data pipelines, AI services and intelligent applications across the modern data center."

TN85-B8261 with AMD Instinct™ MI350P GPU
MiTAC will also showcase the TN85-B8261, a versatile 2U dual-socket AI server designed to accelerate demanding AI, HPC, and data analytics workloads. Featuring support for AMD Instinct™ MI350P GPUs and AMD Pensando™ Pollara 400 AI NIC, the platform delivers powerful GPU-accelerated performance for AI training and inference. Equipped with dual AMD EPYC™ 9005/9004 Series processors, up to 6TB of DDR5 memory, and support for four double-width PCIe 5.0 GPUs, the TN85-B8261 provides the compute density and scalability required by modern AI data centers. Eight hot-swap NVMe U.2 bays and Titanium-efficiency redundant power supplies further enhance performance, reliability, and deployment flexibility. For more details, please visit MiTAC's dedicated page for AMD Instinct™ MI350P GPU solutions.

52U High-Density AI Liquid-Cooled Rack: G4826Z5 AI servers with up to 96 AMD Instinct™ MI355X GPUs and AMD Pensando™ Pollara 400 AI NIC
MiTAC will also present a rack-scale AI infrastructure solution engineered for next-generation AI factories and large-scale model training. Integrating 12 MiTAC G4826Z5 AI servers with up to 96 AMD Instinct™ MI355X GPUs, the platform delivers 50% higher GPU density per rack compared to conventional AI deployments while reducing data center footprint by up to 33%. Built on MiTAC's turnkey rack-scale integration approach, the solution combines advanced liquid cooling, optimized power delivery, networking, and CDU integration to maximize performance, efficiency, and scalability, enabling customers to deploy high-density AI infrastructure with lower TCO and accelerated time-to-deployment.

G8825Z5 with Diamond Cooling^® technology
MiTAC will also showcase the G8825Z5 Diamond-cooled AI server, an advanced platform designed to maximize AI performance and energy efficiency for next-generation data centers. Leveraging Akash Systems' proprietary Diamond Cooling^® technology and powered by AMD Instinct™ MI350X GPUs and AMD Pensando™ Pollara 400 AI NIC, the solution delivers up to 50% more AI tokens per watt under standard operating conditions while maintaining throttle-free performance at inlet temperatures above 95°F. The innovative thermal design reduces cooling requirements and improves overall system efficiency, helping lower both CapEx and OpEx. By combining high-performance AI acceleration with breakthrough cooling innovation, the G8825Z5 enables customers to deploy more sustainable and cost-effective AI infrastructure for large-scale training and inference workloads.

AI Rack Testing Capability
MiTAC also offers three AI racks powered by AMD Instinct™ MI350 GPUs and AMD EPYC™ CPUs, enabling up to 96 engineers and developers to remotely log in for testing and validation. This shared-access environment supports collaborative AI development, workload benchmarking, and faster iteration for teams evaluating next-generation AI infrastructure.

Empowering Future-Ready Enterprise Transformation

MiTAC delivers diversified AI infrastructure, spanning CPU systems to liquid-cooled racks. Combining AMD platforms with MiTAC's integration expertise seamlessly optimizes customer workloads. This builds future-ready infrastructure for AI, HPC, and enterprise transformation.

Please visit the 6^th Gen EPYC Server CPUs landing page to learn more about the launch details. Or see MiTAC Computing Technology Corporation website: https://www.mitaccomputing.com/

AMD, the AMD Arrow logo, EPYC, AMD Instinct, and the combination thereof are trademarks of Advanced Micro Devices, Inc.

SOURCE MiTAC Computing Technology Corp.

MiTAC Computing Technology Corporation, Louise Chang, [email protected]