At the COMPUTEX 2024 exhibition, ASUS showcased a wide range of AI servers, from productive 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 such as the NVIDIA MGX-based ESC NM1-E1, ESC NM2-E1, and ESR1-511N-M1, along with NVIDIA HGX H200 and Blackwell GPU-based solutions, were also on display.
ASUS Showcases New Servers at COMPUTEX 2024!
ASUS is focused on delivering end-to-end solutions and software services ranging from hybrid servers to edge computing deployments. At this event, ASUS demonstrated its expertise in server infrastructure, storage, data center architecture, and secure software platforms, aiming to accelerate the widespread adoption of AI across different industries.
The star of the show, ASUS ESC AI POD, takes the collaboration with NVIDIA to the next level by developing high-performance and efficient AI servers. Featuring the NVIDIA Blackwell-supported scalable large-form-factor system, ESC AI POD enables fast and direct communication for training and real-time inference operations of large language models (LLMs) with trillions of parameters.
Equipped with the latest NVIDIA GB200 Grace Blackwell Superchip and fifth-generation NVIDIA NVLink technology, the system delivers optimal AI processing performance, compatible with both liquid-air and liquid-liquid cooling solutions.
Paul Ju, ASUS Corporate Vice President and Co-Head of the Open Platform Business Group, emphasized the synergy between ASUS and NVIDIA in AI processing. “Our collaboration with NVIDIA, a global leader in AI processing, supports our expertise. Together, we develop AI servers with superior performance and efficiency,” he stated.
To support the rise of productive AI applications, ASUS introduced a comprehensive server series based on the NVIDIA MGX architecture. This includes the 2U ESC NM1-E1 and ESC NM2-E1 (NVIDIA GB200 NVL2 servers) and the 1U ESR1-511N-M1.
Designed to serve large-scale AI and HPC applications, ASUS’ latest HGX servers feature the 5th Generation Intel Xeon Scalable processors and NVIDIA Blackwell Tensor Core GPUs (ESC N8) and AMD EPYC 9004 processors and NVIDIA Blackwell GPUs (ESC N8A).
These servers leverage advanced thermal solutions for optimal performance and lower power consumption. Tailored for AI and data science advancements, these powerful NVIDIA HGX servers offer GPU-to-NIC configurations for maximum efficiency in tasks requiring intensive processing.
Going beyond providing top-tier hardware, ASUS specializes in developing custom data center solutions tailored to meet the needs of businesses. From system validation to remote deployment, their services ensure seamless operations crucial for accelerating AI development. Additionally, ASUS AI server solutions with integrated NVIDIA BlueField-3 SuperNICs and DPUs support the NVIDIA Spectrum-X Ethernet network platform, delivering the best network capabilities for productive AI infrastructures.
Conclusion
ASUS’s unveiling of AI servers at COMPUTEX 2024 demonstrates its commitment to advancing AI technology across various sectors. Through strategic partnerships and innovative solutions, ASUS continues to push the boundaries of AI processing, enabling businesses to harness the power of artificial intelligence for enhanced productivity and efficiency.
