ASUS, together with its subsidiary Taiwan Web Service Corporation (TWSC), will unveil the GenAI POD solution, its holistic approach, to meet the growing demand for AI supercomputers at ISC 2024. ü announced. At the forefront of the AI revolution, ASUS offers NVIDIA MGX models, including the ESC NM1-E1 and ESR1-511N-M1 models, as well as the ESC N8A-E12 and RS720QN-E11-RS24U HGX GPU servers. exhibited its supported artificial intelligence servers. These solutions are supported by TWSC’s proprietary resource management platform and software stacks, enabling a variety of generative AI and large language model (LLM) training jobs. job It allows you to easily overcome your burdens. Featuring innovative thermal designs, these integrated solutions provide comprehensive data center interiors with robust software platforms to enable customers to succeed in their AI efforts. à It can be adapted to appeal to businesses by providing solutions.
ASUS NVIDIA MGX servers: AI solution tailored to meet specific needs
NVIDIA MGX supported ASUSA ESC NM1-E1 is powered by NVIDIA GH200 Grace Hopper Superchip. This powerful combination ensures exceptional performance and efficiency with 72 Arm Neoverse V9 CPU cores and NVIDIA NVLink-C2C technology, enabling AI-driven data centers, high-performance computing (HPC), data analytics and NVIDIA Omniverse apps It offers an ideal choice for users and promises transformative improvements in performance and memory capacity.
ASUS ESR1-511N-M1 server, another product highlighted by ASUS, uses the power of NVIDIA GH200 Grace Hopper Superchip. THIS server is designed to address large-scale AI and HPC applications by facilitating deep learning (DL) training and inference, data analytics, and high-performance computing. designed. The ESR1-511N-M1 features an improved thermal solution for optimal performance and lower power utilization efficiency ratio (PUE) to comply with ESG trends. Flexible configuration featuring 1U design with the highest compute density and support for up to four E1.S native drives via NVIDIA BlueField-3-3 ±river, three PCI Express When combined with the (PCIe) 5.0 x16 slot, it facilitates smooth and fast data transfer.
ASUS NVIDIA HGX servers: End-to-end H100 takes AI to the next level with the power of eight GPUs
Designed for productive AI with optimized servers, data center infrastructure and AI software development capabilities, ASUS ESC N8A-E12 is powered by dual AMD EPYC 9004 processors and eight NVIDIA H100 Tensor Core GPUs. A rugged 7U dual-socket server that packs its punch. The product features an advanced thermal solution to deliver optimum performance and lower PUE rate. Designed for AI and data science advancements, this powerful HGX server offers a unique single GPU single NIC configuration for maximum throughput on compute-intensive tasks.
Designed for high-performance and compute-intensive workloads, ASUSÂ RS720QN-E11-RS24U features a high-density server featuring NVIDIA Grace CPU Superchip with NVIDIA NVLink-C2C technology. This innovative solution features four hubs in a 2U4N chassis to deliver PCIe 5.0 compatibility and outstanding performance for dual-socket CPUs RS720QN comes with a compact design that can -Making the E11-RS24U ideal for data centers, web servers, virtualization clouds and hyperscale environments.
ASUS D2C cooling solution
Direct-to-chip (D2C) cooling provides faster deployment thanks to lower PUE as well as a fast and simple solution that takes advantage of existing infrastructure ± can be done It is provided. ASUSÂ RS720QN-E11-RS24U supports manifolds and cooling plates, enabling various cooling solutions. Additionally, ASUS servers feature a backdoor heat exchanger compatible with standard rack server designs. Thus, there is no need to replace all the shelves, and only the back door is sufficient to activate liquid cooling in the shelves. Working closely with industry-leading cooling solution providers, ASUS delivers comprehensive enterprise-level cooling solutions and reduces data center PUE, carbon emissions and greener by minimizing energy consumption is committed to helping with the design and construction of data centers.
Productive AI POD solutions
TWSC is an NVIDIA Partner with the National Center for High Performance Computing’s (NCHC) TAIWANIA-2 (#10 / Green 500, November 2018) and FORERUNNER 1 (#92 / Green 500, November 2023) supercomputer series Highly experienced in delivering large-scale AIHPC infrastructure deployment and joint maintenance for network cloud partnership (NCP).
Additionally, TWSC’s AI Foundry Service empowers users by enabling rapid deployment of AI supercomputers and flexible model optimization for AI 2.0 applications. It allows users to tailor their demands regarding artificial intelligence specifically to themselves.
TWSC’s productive AI Platform On Demand (POD) solutions are widely used by providing enterprise-level AI infrastructure with rapid deployment and comprehensive end-to-end services. SMS and cyber security standards are met. Our solutions enable success stories to be written in academies, research and medical institutions. Comprehensive cost management options optimize power consumption and facilitate OPEX. This makes TWSC technologies an attractive choice for organizations looking for a reliable and sustainable productive AI platform.
