On September 24, the CCF HPC China 2024 event took place grandly at the China Optics Valley Convention & Exhibition Center in Wuhan. The event gathered hundreds of leading scholars from the computing power field and nearly a hundred top companies globally. It focused on cutting-edge topics like AI integration, computing power innovation, and large-scale model applications. Gooxi showcased its cutting-edge technology and solutions, exploring new opportunities and challenges in high-performance computing (HPC) in the intelligent computing era.

HPC (High-Performance Computing) has always been a frontier technology, demanding the highest standards in hardware infrastructure. Gooxi, a leading provider of computing power solutions, offers a range of server architectures to meet the specific requirements of HPC clusters. At this event, Gooxi displayed its top-tier HPC server lineup, known for its computing power, high bandwidth, and optimal performance in data handling.

1. Intel Eagle Stream Platform 4U 8-GPU AI Server

Unleashing the Full Potential of High-Performance Computing

Gooxi's Intel Eagle Stream platform 4U 8-GPU AI server is a powerhouse for high-performance tasks. With robust computational abilities, high-bandwidth structures, and flexible GPU topologies, it supports one or two 4th/5th generation Intel® Xeon® scalable processors. The server is configured to support direct CPU-GPU communication and GPU switch expansion, meeting various AI workloads. It provides up to 8 GPUs in a 4U space, making it a perfect solution for AI cloud, AI training, and HPC applications, with flexible support for SATA/SAS/NVMe drives.

2. AMD Milan Dual-Processor 4U 10-GPU AI Server

A Balance of High Performance and Cost Efficiency

The AMD Milan dual-processor 4U 10-GPU AI server combines excellent computing power, lower cost, and extreme scalability, making it an ideal choice for AI in the HPC sector. Powered by one or two AMD EPYC 7003/7002/7001 processors, it offers enhanced memory bandwidth and resolves performance bottlenecks in HPC environments. This server supports up to 10 double-width GPUs, interlinked via GPU switches, and provides up to 20 PCIe slots for further expansion. It is well-suited for deep learning, training inference, big data analysis, and scientific computing.

3. Gooxi HPC Cluster System

Unleashing Unlimited Computing Power

As organizations look for powerful AI and big data analysis solutions, high-performance computing infrastructure offers the needed advantages. However, designing, deploying, and managing HPC systems can be daunting. Gooxi continuously explores AI and HPC innovations, aiming to deliver efficient, intelligent, and customizable solutions. These solutions address the industry's computing, storage, and intelligent decision-making demands.

Gooxi's HPC management platform integrates hardware and software for optimal system management, centralizing resource allocation and job scheduling. It also offers real-time monitoring of CPU/GPU usage and system health. This platform can be applied in fields like CAE design simulation, physical chemistry, and AI deep learning.

In addition to its technical prowess, Gooxi offers customization services to tailor HPC deployments to specific customer needs. Optimized for performance, these solutions ensure stability and efficiency during operation, supporting clients' critical workloads.

With the nation's focus on fostering next-gen productivity, Gooxi is committed to working with its partners to create agile, AI-driven computing infrastructure. Together, they aim to push forward the intelligent computing industry and contribute to the rise of high-quality productivity.