ASUS announced a comprehensive range of AI servers at Computex, featuring the debut of the advanced ASUS ESC AI POD with the NVIDIA GB200 NVL72 system. This event highlights the apex of ASUS's AI solutions. The display included the latest NVIDIA MGX-powered systems like ESC NM1-E1 and ESC NM2-E1, and systems fitted with NVIDIA GH200 Grace Hopper Superchips. These technologies cater to a variety of AI applications from generative AI to high-performance computing (HPC).
ASUS's exhibition demonstrated its proficiency in server infrastructure, showcasing its commitment to providing all-encompassing solutions, from hardware to software, tailored for enterprises. The showcase underscored ASUS's intent to propel AI adoption across various sectors through robust server infrastructure and secure software platforms.
Introducing the ASUS ESC AI POD
The ASUS ESC AI POD, which integrates NVIDIA GB200 NVL72 and NVIDIA GB200 Grace Blackwell Superchips, was the spotlight of the showcase. This system presents a large form factor that significantly escalates AI capabilities, equipped with an array of GPUs, CPUs, and switches that enable rapid communication. This configuration is designed to dramatically accelerate large language model (LLM) training and real-time inference. It also supports advanced cooling solutions to sustain optimal performance levels.
Paul Ju, Corporate Vice President for ASUS and Co-Head of Open Platform BG, shared his thoughts on their collaboration with NVIDIA, stating, “Our partnership with NVIDIA, a global leader in AI computing, underpins our expertise. Together, we've harnessed a potent synergy, enabling us to craft AI servers with unparalleled performance and efficiency.”
Kaustubh Sanghani, vice president of GPU product management at NVIDIA, also spoke on the collaboration, “ASUS has strong expertise in server infrastructure and data center architecture, and through our work together, we're able to deliver cutting-edge accelerated computing systems to customers. At Computex, ASUS will have on display a wide range of servers powered by NVIDIA's AI platform that are able to handle the unique demands of enterprises.”
Advancing the boundaries of AI and HPC
ASUS also presented its MGX and HGX scalable solutions designed to meet the high demands of generative AI and HPC. The latest solutions include servers like ESC N8 and ESC N8A, equipped with the latest NVIDIA Blackwell Tensor Core GPUs and enhanced thermal solutions to ensure peak performance.
Moreover, ASUS excels in offering software-defined data centre solutions that go beyond conventional hardware offerings, incorporating comprehensive software solutions designed to facilitate seamless AI operations. These solutions are crucial for businesses looking to enhance their AI development capabilities. ASUS's tailored services range from system verification to remote deployment, ensuring their systems are optimised for enterprise needs.
Additionally, ASUS integrates NVIDIA BlueField-3 SuperNICs and DPUs in its AI servers to support the NVIDIA Spectrum-X Ethernet networking platform, providing exceptional networking capabilities essential for generative AI infrastructures.
