Press Release

Taipei, Taiwan – 12, November 2025 – Inventec, a global leader in AI and data center solutions, will showcase its next-generation computing innovations at SuperComputing 2025 (SC25), held from November 18 to 20 at the America’s Center Convention Complex in St. Louis, Missouri. This year, Inventec redefines the exhibition experience by presenting its entire lineup through interactive 3D virtual models, offering visitors an immersive way to explore the company’s latest advancements in AI and HPC infrastructure.

As AI continues to transform industries worldwide, Inventec remains committed to building powerful, scalable, and energy-efficient computing systems. At SC25, the company will unveil a comprehensive portfolio designed for AI training, inference, and high-performance computing (HPC), empowering customers to meet the accelerating computational demands of the AI era.

At the center of Inventec’s showcase is the Artemis II rack solution, built on NVIDIA GB300 NVL72 systems and based on the NVIDIA MGX platform, Artemis II is engineered to support large-scale generative AI workloads that demand exceptional compute density, energy efficiency, and reliability. Alongside Artemis II, Inventec will present the P9000IG7 (LC), P9000IG7 (AC), and P9000AG7 (AC), a family of high-performance AI servers ready for rack-level deployment. These systems integrate NVIDIA Blackwell and NVIDIA Blackwell Ultra GPUs with Intel® Xeon® 6 processors and AMD EPYC™ 9005 Series processors, offering flexible scalability and optimized thermal management in both liquid- and air-cooled configurations. Expanding its lineup, Inventec introduces the P4200G7, a 4U NVIDIA MGX-based server equipped with NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and NVIDIA ConnectX-8 SuperNIC, delivering outstanding visualization and compute performance for enterprise AI and digital content creation.

Further strengthening its HPC server portfolio, Inventec introduces the K888G7, C805G7, and C805DG7, purpose-built for scientific research, engineering simulations, and other compute-intensive workloads. The K888G7, powered by dual Intel® Xeon® 6 processors with either P-cores or E-cores architecture, supports up to four double-width GPUs and provides exceptional flexibility with both air and liquid cooling. The C805G7 and C805DG7, built on 5th Gen AMD EPYC™ processors, extend Inventec’s high-density compute capabilities for modeling, simulation, and high-throughput computing. The C805DG7’s 2U dual-node design enables independent node operation, optimizing rack space and ensuring reliability for clustered and containerized HPC environments. Together, these systems demonstrate Inventec’s commitment to delivering modular, high-performance infrastructure that drives the next generation of scientific and technological innovation.

Visitors to the Inventec booth at SC25 will be able to experience all products in a fully interactive 3D virtual environment, exploring detailed design features and real-time system interactions. This innovative approach reflects Inventec’s vision to merge technology and creativity, delivering not only cutting-edge hardware but also a new way to experience the future of AI infrastructure.