ASUS showcases NVIDIA GB300 NVL72 at SC25
ASUS presented its Artificial Intelligence (AI) infrastructure portfolio utilizing NVIDIA Blackwell architecture at Supercomputing 2025 (SC25). The offerings included solutions aimed at enhancing computing performance for AI and High performance computing (HPC).
The ASUS XA GB721-E2 system featured liquid cooling and integrated networking, delivering 10-petaflops-class computing power. The newly introduced ESC8000A-E13X supported enterprise acceleration with eight NVIDIA RTX PRO 6000 GPUs and a 400G SuperNIC to facilitate large-model AI and HPC workloads.
ASUS highlighted its commitment to providing an end-to-end AI infrastructure by combining advanced cooling and professional services. This effort aims to democratize powerful AI technologies, ensuring that a wide range of clients have access to scalable solutions.
The XA GB721-E2 serves as a rack-scale AI system, using NVIDIA’s GB300 NVL72 to facilitate the deployment of energy-efficient computing for national cloud workloads. Additionally, the Embedded Security Controller (ESC) NM2N721-E1, also powered by NVIDIA technology, enhances capabilities in sovereign AI initiatives.
At the event, ASUS introduced the ESC8000A-E13X, designed for large-model training and inference, ensuring optimal performance with cooling and connectivity features. This system underscores the emphasis on compute density and efficiency across a spectrum of AI applications.
ASUS’ comprehensive lineup extends from large data centers to personal computing systems, emphasizing flexibility and implementation ease. The lineup includes systems that support multi-GPU configurations and AI model training, catering to enterprise requirements and research entities.
The partnership with Vertiv was also featured, focusing on rack-scale deployment solutions for AI infrastructures, enhanced with advanced cooling capabilities. This collaboration is intended to improve energy efficiency and facilitate industrial-scale AI deployments.
ASUS continues to highlight its AI and supercomputing infrastructure through strategic partnerships, advanced engineering, and deployment support to ensure reliability and scalability for clients’ AI projects.