MiTAC Computing Showcases Full-Stack Innovation to Scale Open AI Infrastructure at OCP Global Summit 2026
SAN JOSE, Calif. — October 12, 2026 — MiTAC Computing Technology
Corporation, a subsidiary of MiTAC Holdings Corporation (TWSE: 3706) and a
global leader in high-performance server solutions, today unveiled its latest
AI infrastructure innovations at the 2026 OCP Global Summit, Booth D22,
under the theme “Scaling Open AI Infra with Full-Stack Innovation.”
Large-scale AI clusters and AI factories demand
more than compute performance alone. MiTAC Computing takes a full-stack
approach, transforming fragmented infrastructure into a cohesive, scalable
foundation designed for deployment, management, and expansion.
From
Scalable AI Factories to Modular Data Centers: Flexible Infrastructure for
Every Scale
As AI workloads expand from training, inference,
and RAG to Agentic AI, infrastructure needs vary by scale and deployment
environment. MiTAC Computing’s AI Infrastructure Stack spans compute,
storage, and liquid-cooled servers through open platform firmware, operating
systems, AI software ecosystems, and AI Pod management, providing a consistent,
scalable foundation for diverse workloads.
For large-scale, centralized AI environments, MiTAC
Computing delivers its Scalable AI Factory, combining full-stack
capabilities with rack-scale integration to support high-density computing and
continued expansion. For distributed sites or rapid buildouts, MiTAC offers
modular data center solutions. The TonoForge™ Modular Data Center, developed with Tonomia, incorporates MiTAC G4826Z5 liquid-cooled AI servers and rack-scale integration. Its
proven 12-week deployment demonstrates how the modular approach can
accelerate the deployment of high-performance GPU AI clusters.
From centralized AI factories to distributed
modular data centers, MiTAC gives enterprises the flexibility to build AI
infrastructure around their compute requirements, site conditions, and
deployment needs.
Breaking
Thermal Barriers with Advanced Liquid Cooling
Higher compute density and rack-scale AI factories
are making advanced cooling critical to unlocking greater compute performance.
MiTAC Computing is showcasing multiple advanced liquid cooling solutions,
including a PoC with ZutaCore that integrates the MiTAC
G4826Z5 with ZutaCore® HyperCool® Waterless
Two-Phase Liquid Cooling technology. Using a dielectric fluid that changes
phase from liquid to vapor directly at the chip, the technology removes heat
generated by CPUs and GPUs at the source.
The technology demonstrates the potential to
support more than 4,000W per chip and average heat flux exceeding 250W/cm².
Under defined operating conditions, an annualized model indicates a pPUE as
low as 1.03, highlighting the value of two-phase cooling for increasingly
power-dense AI infrastructure.
On October 14, MiTAC Computing will present an Expo
Hall Session exploring the evolution of high-density AI systems toward
next-generation rack-scale platforms and how advanced liquid cooling can
overcome thermal constraints to unlock greater compute performance.
Extending
Openness Across the AI Stack
Alongside advances in hardware performance, MiTAC
Computing continues to develop its Open Source Firmware Solution. The G4826Z5
demonstrates full-stack integration from bare metal to AI applications. At the
platform level, MiTAC’s internally developed MiOPF™ supports fast system
initialization, while MiOBMC™ provides real-time hardware monitoring and
telemetry, giving customers greater firmware control and customization
flexibility while reducing dependence on a single vendor.
Across the software ecosystem, the G4826Z5 runs Canonical-certified
Ubuntu 26.04 OS and integrates Kubernetes and the DDN Data
Intelligence Platform. It also incorporates AMD ROCm™ open software stack to unlock the compute
performance of AMD Instinct™ MI355X GPUs. MiTAC Computing is working with
Rafay on GPU cluster management to support flexible GPU-as-a-Service
(GPUaaS) resource pools, enabling on-demand compute allocation based on AI
workload requirements while simplifying deployment and management.
In an Executive Session later today, MiTAC
Computing will explore open infrastructure management through technologies such
as Open Platform Firmware (OPF), OpenBMC, and Redfish. The discussion will
examine how open standards and OCP community collaboration can improve infrastructure
interoperability, and how MiCoreView™, MiTAC’s infrastructure management
software built on open-source technologies, can further extend management
capabilities from the platform level to rack and cluster scale. MiTAC will also
share practical insights into deployment, telemetry, data center rack and
cluster management, and system orchestration.
New
Platforms Target OCP Computing and AI Inference
MiTAC Computing is making the global debut of the
C2611Z6 and showcasing the latest G4520G6E3BC-N configuration,
expanding its portfolio for high-density OCP computing and enterprise AI
inference.
- C2611Z6: A liquid-cooled multi-node server designed for OCP ORv3 racks and powered by the latest AMD EPYC™ 9006 Series Server processors. It integrates four compute nodes in a compact 2OU form factor, with each node supporting 16 DDR5-8000 RDIMM slots and OCP 3.0 expansion for high-density, liquid-cooled infrastructure.
- G4520G6(G4520G6E3BC-N): l
A 4U2S GPU server built for enterprise AI
inference. Powered by Intel® Xeon® 6
processors, the system supports up to eight NVIDIA RTX PRO 4500 / NVIDIA 6000 Blackwell Server Edition GPUs or NVIDIA H200 GPUs. The configuration also features
12 NVMe E3.S drive bays, combining high-density GPU compute and
high-speed storage in a single 4U platform for data-intensive inference
workloads.
MiTAC Computing invites OCP Global Summit attendees to visit Booth D22 and experience how full-stack innovation is helping scale open AI infrastructure.
Learn more:
######
About MiTAC Computing Technology Corporation
MiTAC Computing Technology Corp., a subsidiary of
MiTAC Holdings, delivers comprehensive, energy-efficient server solutions
backed by industry expertise dating back to the 1990s. Specializing in AI, HPC,
cloud, and edge computing, MiTAC Computing employs rigorous methodologies to
ensure uncompromising quality—across barebones, systems, racks, and cluster
levels—fully achieving performance and integration. This commitment to quality
at every level sets MiTAC Computing apart in the industry.
With a worldwide presence and end-to-end
capabilities—from R&D and manufacturing to global support—MiTAC Computing
provides agile, customized platforms for hyperscale data centers, HPC, and AI
applications, ensuring optimal performance and scalability to meet unique
business needs. By leveraging the latest advancements in AI and liquid cooling,
and unifying the MiTAC brand with Intel DSG and TYAN server products, MiTAC
Computing stands out for its innovative, efficient, and reliable server
technology as well as its hardware and software integrated solutions—empowering
businesses to meet future challenges.
MiTAC Computing Technology Corporation
website: https://www.mitaccomputing.com/
AMD, the AMD logo, EPYCTM, AMD InstinctTM, and combinations thereof are trademarks of Advanced Micro Devices, Inc.
