ASUS has announced its participation in Supercomputing Asia 2025 (SCA25), set to take place from 10 to 13 March at the Marina Bay Sands Expo and Convention Centre, Singapore. Visitors can explore ASUSโ next-generation high-performance computing (HPC) solutions at Booth B3, where the company will unveil its latest range of AI servers and liquid-cooled rack solutions, designed to meet a variety of workloads.
As industries undergo rapid digital transformation, ASUS aims to provide cutting-edge GPU and rack servers to optimise AI applications, machine learning, cloud computing, and enterprise data management.
AI-optimised GPU servers for generative AI and machine learning
ESC 8000A-E13P: Built for large-scale AI workloads – Leading the showcase is the ESC 8000A-E13P, a high-performance computing (HPC) server that supports NVIDIA MGX modular architecture. This system is built for large-scale AI deployments and features high-density GPU configurations for a variety of demanding applications. It supports up to eight NVIDIA H200 NVL dual-slot cards, each with a 600-watt power rating. To enhance data transfer speeds and system performance, the ESC 8000A-E13P is equipped with five PCIe 5.0 slots, optimised for NVIDIA BlueField-3 and ConnectX-8 SuperNICs. These components improve east-west traffic handling, while NVIDIA BlueField-3 DPUs enable high-efficiency software-defined networking, storage, and cybersecurity processing.
ESC N8-E11V: Advanced performance with NVIDIA HGX H200 – For generative AI workloads, ASUS will showcase the ESC N8-E11V, powered by the NVIDIA HGX H200 platform. This AI server is designed for data-intensive tasks, including AI, machine learning, and HPC applications. A key feature of the ESC N8-E11V is its focus on power efficiency and thermal optimisation. It supports both air-cooling and direct-to-chip (D2C) liquid-cooling solutions. With dedicated airflow tunnels for CPUs and GPUs, heat is efficiently expelled, ensuring the system remains cool and energy-efficient even under heavy workloads.
High-performance rack servers for cloud and HPC
RS520QA-E13-RS8U: Optimised for HPC and enterprise workloads – The RS520QA-E13-RS8U is a high-efficiency multi-node server designed for HPC, electronic design automation (EDA), and cloud computing. Powered by AMD EPYC 9005 processors, it supports 20 DIMM slots per node, ensuring scalability and memory expansion. With support for PCIe 5.0 and OCP 3.0, this server delivers enhanced performance and flexibility for handling data-heavy tasks in research, enterprise, and cloud applications.
RS720A-E12-RS24U: Scalable 2U server for AI and data processing – The RS720A-E12-RS24U is a next-generation 2U rack server designed for accelerated computing. Its parallel CPU and memory layout enables efficient airflow, improving thermal performance for AI applications and storage-intensive workloads.
RS920Q-E12-RS16: Intel-powered density and performance for HPC – Designed for intensive HPC workloads, the RS920Q-E12-RS16 features Intel Xeon 6 processors, offering a high-density, high-performance solution. It incorporates direct-to-chip liquid cooling, ensuring optimal thermal management in demanding HPC and cloud computing environments.
Hybrid storage server for SMB applications
VS320D-RS26: Efficient storage with a reduced footprint – For businesses needing cost-effective, high-capacity storage, ASUS is introducing the VS320D-RS26. This 2.5-inch, 2U 26-bay hybrid flash storage system provides 8% more capacity than standard 2U 24-bay products, reducing cost per terabyte and minimising IT rack space. It is ideal for enterprise applications such as virtualisation, databases, mail servers, and virtual desktop infrastructure (VDI).
Exclusive AI innovations and industry insights
In addition to its server lineup, ASUS will feature SoftServeโs Jade, a Generative AI-powered virtual assistant, offering human-like interactions and personalised responses for enhanced customer service.
ASUS has also confirmed keynote sessions featuring industry leaders, including Peter Wu, CEO of TWSC, Taiwan AI Cloud, and Dr Robby Tan, Chief Scientist of ASUS Intelligent Cloud Services (AICS). They will share insights on AI advancements and supercomputing.
ASUS is also launching a proof-of-concept (POC) initiative for the ASUS AI POD, an NVIDIA Blackwell architecture-powered computing solution. A limited group of innovators will have the opportunity to participate in this exclusive trial. Interested applicants can complete a survey to be guided through the next steps by NVIDIA GB200 Grace Blackwell platform specialists.
