ASUS and INNO3D unveil new NVIDIA MGX-based AI servers

ASUS and INNO3D have introduced new high-performance server solutions designed to accelerate AI workloads, including large language models (LLMs), reinforcement learning, and high-performance computing (HPC).

At the OCP Global Summit 2026, ASUS unveiled the XA P2N-E2, a 2U server based on NVIDIA MGX architecture. Powered by NVIDIA Vera CPUs, the system is designed for agentic AI tasks such as code execution, data processing, and orchestration. It supports up to two dual-slot NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. ASUS also showcased the XA VR721-E3 AI POD, which utilizes a liquid-cooled architecture to integrate 36 NVIDIA Vera CPUs and 72 NVIDIA Rubin GPUs.

INNO3D announced the AGS-4UMGX-R1, a 4U GPU server also built on the NVIDIA MGX platform. This system features dual Intel Xeon 6 processors and supports up to eight NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. To facilitate ultra-fast networking, the server includes the NVIDIA ConnectX-8 SuperNIC with eight 400Gbps QSFP ports, targeting enterprise environments for digital twins and AI training.