Unlocking a New AI Chapter: INFINITIX, MiTAC, and GRAID Showcase Software-Hardware Integration at INTEROP Japan

As generative AI rapidly proliferates globally, major enterprises are successively deploying computing power centers, and hardware resources are becoming increasingly sought after in the market. However, enterprises still face numerous technical bottlenecks and management hurdles in the process of adopting AI applications. The optimal solution for computing resource management, and Taiwan's leading brand in computing power management—INFINITIX—helps enterprises significantly boost the utilization of hardware such as GPUs and servers through its core product: the comprehensive AI management platform, AI-Stack.

Today (11th), INFINITIX, in collaboration with MiTAC and Graid, is participating in INTEROP JAPAN, one of the world's renowned and iconic IT and network communication solution exhibitions, and Japan's largest IT show. We look forward to leveraging this opportunity to expand into international markets and create more global business cooperation opportunities.

INFINITIX CEO Wen Yu Chen stated that in recent years, INFINITIX has been actively expanding its presence in the Asia-Pacific market, establishing new offices in Japan and South Korea, and recently strengthening its sales channels in Southeast Asia. He noted the high level of interest from Japanese and Korean clients during the recent COMPUTEX exhibition. Chen looks forward to leveraging the excellent opportunity at INTEROP Japan to exhibit alongside top hardware partners like MiTAC Japan and Graid. This collaboration will integrate software, servers, and storage into highly efficient solutions, creating more potential business opportunities.

Chen emphasized that INFINITIX's core product, AI-Stack, is highly competitive and stands as a globally leading computing power management platform. Its crucial GPU partitioning technology allows a single GPU to stably handle multiple parallel tasks and can also integrate multiple GPUs for efficient cross-node computing. Chen further explained that AI-Stack provides efficient GPU management and Machine Learning Operations (MLOps) functionalities. It not only supports integration with various popular development tools but also offers resource isolation, permission management, and quota limits. Additionally, it can automatically execute single or batch tasks, helping teams manage GPU resources more precisely and efficiently. This enables users to build a container-friendly environment, making it easy to select desired GPU resources and AI development frameworks. From model creation and training to inference, AI-Stack provides a complete infrastructure management solution, helping enterprises accelerate AI application deployment.

MiTAC Computing, a subsidiary of MiTAC Holdings, has delivered comprehensive, energy-efficient server solutions since the 1990s. With a global presence and end-to-end capabilities, from R&D and manufacturing to global support, MiTAC Computing provides tailored platforms for hyperscale data centers, HPC, and AI applications, guaranteeing optimal performance and scalability. Featured at Interop Tokyo 2025 are MiTAC’s latest AI and HPC servers, liquid cooling solutions, and OCP full rack level integration capabilities, where true performance and integration matter most.

Graid Technology, a Silicon Valley–based innovator, will highlight its patented GPU-based RAID solution, SupremeRAID™, the world’s fastest and most secure NVMe/NVMeoF RAID card. Unlike traditional RAID, SupremeRAID™ leverages GPU acceleration to eliminate bottlenecks and dramatically increase data throughput. At the event, Graid will showcase the SupremeRAID™ SR-1010 integrated into the MiTAC B8261 platform. This powerful HPC storage system, model B8261T85E24HR-2T, supports 24 NVMe SSDs and is ideally suited for data-intensive environments such as Synchrotron Radiation research, high-end analytics, and enterprise AI pipelines.