ASUS showcased a wide range of AI servers at COMPUTEX 2024, from generative AI to innovative storage solutions. The highlight of the event was the new ASUS ESC AI POD, featuring the NVIDIA GB200 NVL72 system. Additionally, servers based on NVIDIA MGX such as ESC NM1-E1, ESC NM2-E1, and ESR1-511N-M1, as well as solutions with NVIDIA HGX H200 and Blackwell GPUs, were also displayed.

ASUS presents new AI servers at COMPUTEX 2024!

ASUS focuses on providing end-to-end solutions and software services, from hybrid servers to edge computing deployments. At this event, the company showcased its expertise in server infrastructure, storage, data center architecture, and secure software platforms, aiming to promote the widespread adoption of AI across various sectors.

The star of the event, ASUS ESC AI POD, takes ASUS’s expertise in developing high-performance and efficient AI servers one step further through collaboration with NVIDIA. The scalable, large form-factor system supported by NVIDIA Blackwell is designed for training and real-time inference of large language models (LLMs) with trillions of parameters. The system integrates the latest NVIDIA GB200 Grace Blackwell Superchip and fifth-generation NVIDIA NVLink technology, offering optimal AI processing performance with both liquid-air and liquid-liquid cooling solutions.

Paul Ju, ASUS Corporate Vice President and Co-President of Open Platform BG, stated, “Our collaboration with NVIDIA, a world leader in AI processing, supports our expertise. Together, we create a unique synergy, enabling us to develop AI servers with superior performance and efficiency.”

To support the rise of generative AI applications, ASUS introduced a complete series of servers based on NVIDIA MGX architecture, including the 2U ESC NM1-E1 and ESC NM2-E1 (NVIDIA GB200 NVL2 servers) and the 1U ESR1-511N-M1.

ASUS NVIDIA MGX-based products, powered by the NVIDIA GH200 Grace Hopper Superchip, are designed to serve large-scale AI and HPC applications, providing seamless and fast data transfers, deep learning (DL) training and inference, data analytics, and high-performance computing.

Among ASUS’s latest HGX servers designed for high-performance computing (HPC) and AI are the ESC N8, supported by 5th Gen Intel Xeon Scalable processors and NVIDIA Blackwell Tensor Core GPUs, and the ESC N8A, supported by AMD EPYC 9004 processors and NVIDIA Blackwell GPUs. These servers benefit from an advanced thermal solution for optimal performance and lower power consumption. Designed for AI and data science advancements, these powerful NVIDIA HGX servers offer a GPU-to-NIC configuration for maximum efficiency in intensive tasks.

ASUS specializes in developing customized data center solutions that go beyond providing high-end hardware, offering comprehensive software solutions tailored to business needs. Services covering everything from system validation to remote deployment ensure seamless operations essential for accelerating AI development. Additionally, integrated NVIDIA BlueField-3 SuperNICs and DPUs with ASUS AI server solutions support the NVIDIA Spectrum-X Ethernet network platform, providing the best networking capabilities for generative AI infrastructures.