Powering AI innovation with scalable, high-performance infrastructure
- Advanced compute-intensive servers: ASUS scalable server solutions powered by Intel Xeon 6 with DC-MHS modular architecture
- Comprehensive AI & supercomputing solutions: NVIDIA® Blackwell, HGX and MGX, Intel® Xeon® 6, Intel Gaudi® 3 AI accelerator and AMD EPYC™ CPUs
- End-to-end AI infrastructure: ASUS delivers fully integrated hardware, software and cloud-based platforms, offering seamless AI deployment for enterprises
TAIPEI, Taiwan, March 13, 2025 — ASUS today announced its showcase of comprehensive AI infrastructure solutions at CloudFest 2025, bringing together cutting-edge hardware powered by Intel® Xeon® 6 processors, NVIDIA® GPUs and AMD® EPYC™ processors. The company will also highlight its integrated software platforms, reinforcing its position as a total AI solution provider for enterprises seeking seamless AI deployments from edge to cloud.
Intel Xeon 6-based AI solutions and Gaudi 3 Acceleration for generative AI inferencing and fine tuning training
ASUS Intel Xeon 6-based servers leverage the Data Center Modular Hardware System (DC-MHS) architecture, providing unparalleled scalability, cost-efficiency and simplified maintenance. ASUS will showcase a comprehensive Intel Xeon 6 family of processors at CloudFest 2025, including the RS700-E12, RS720Q-E12. and ESC8000-E12P-series servers. The ESC800-E12P-series servers will debut the Intel Gaudi® 3 AI accelerator PCIe® card. This lineup underscores the ASUS commitment to delivering comprehensive AI solutions that integrate cutting-edge hardware with enterprise-grade software platforms for seamless, scalable AI deployments, highlighting Intel’s latest innovations for high-performance AI training, inference, and cloud-native workloads.
ASUS RS700-E12 is a 1U air-cooled server designed for scalable AI operations. Equipped with dual Intel Xeon 6 processors, it supports high-speed DDR5 memory and NVMe storage, delivering exceptional performance for AI inference and training. Its DC-MHS modular design enhances flexibility, allowing data centers to expand or upgrade efficiently while maintaining operational stability.
For more demanding AI tasks, ASUS RS720Q-E12 offers a liquid-cooled 2U4N server solution. This multi-node server ensures superior thermal management with direct-to-chip liquid cooling, making it ideal for HPC and large-scale data processing.
At CloudFest 2025, ASUS will also debut the Intel Gaudi 3 AI accelerator PCIe card alongside ASUS ESC8000-E12P, a 4U server engineered for fine tuning training and inference. The Gaudi 3 PCIe card features a PCIe 5.0 interface, 128GB of HBM2e memory with up to 3.7TB/s bandwidth, and 64 dedicated tensor processor cores optimized for AI workloads. Its built-in AI accelerators and support for mixed-precision operations enhance inferencing and fine-tuning training, while delivering high energy efficiency and scalability. This card integrates seamlessly with ASUS infrastructure, providing a future-ready solution for large-scale AI models, deep learning tasks, and data-intensive machine learning pipelines. Separately, the system powered by dual Intel Xeon 6 CPUs and supports up to eight NVIDIA H200 GPUs, offers unmatched computational power for complex AI workloads.
AI supercomputing excellence for high-performance AI model training and HPC
ASUS will also present its AI supercomputing platforms, led by ASUS ESC N8-E11V, a 7U NVIDIA HGX server engineered for large-scale AI applications. Featuring eight NVIDIA H200 Tensor Core GPUs, dual Intel Xeon 6 processors and advanced liquid cooling, it offers unparalleled performance for training large language models and complex deep learning algorithms. Its modular design ensures easy scalability and efficient maintenance.
The ASUS AI POD with NVIDIA GB200 NVL72 platform stands out as an AI supercomputer designed for the most intensive AI workloads. With 72 NVIDIA Blackwell GPUs, NVLink interconnects, and liquid cooling architectures, it supports trillion-parameter AI model training and inference.
Another highlight is the ASUS RS520QA-E13 server, powered by AMD EPYC 9005 processors, making it an ideal choice for EDA applications or edge computing. This high-density 2U4N server offers flexible memory configurations, and efficient cooling solutions for optimized operation, delivering robust performance to meet the demands of high-density and efficiency-driven workloads.
ASUS as a total AI solutions provider
At CloudFest 2025, ASUS will underscore its role as a total AI solution provider by showcasing its integrated hardware and software solutions. From AI management tools to comprehensive software frameworks, ASUS empowers businesses to deploy, manage, and scale AI applications with ease. By offering end-to-end AI infrastructure solutions, ASUS ensures that enterprises can achieve innovation, efficiency, and growth in the rapidly evolving world of AI.