June 17th, Frankfurt, Germany – Today, Inspur showcased its latest AI HPC products and solutions at ISC19 to help users meet the challenges in the convergence of AI and HPC. These included a variety of innovative high-density, high-performance AI computing platforms, multi-node computing systems and large-scale distributed storage systems, as well as efficient AI HPC management suites and framework optimization solutions.
At present, the trend of AI and HPC convergence becomes more and more obvious. AI has discovered a new application field for HPC, while mature technologies such as multi-machine parallel, high-speed low-latency networks, and scheduling algorithms accumulated in the HPC field greatly reduces the workload of managing and using AI clusters. However, in the actual environment, due to the differences that AI and HPC have in workload, programming model and development application, how to integrate and utilize resources and effectively avoid waste has become a common challenge for AI HPC users.
Inspur is committed to building converged, efficient, flexible and easy-to-use AI HPC infrastructure to help users meet this challenge. Currently, Inspur has a comprehensive AI HPC product line covering computing technologies such as CPU/GPU/FPGA, and is able to meet the different requirements for computing architecture performance and power consumption in various AI HPC application scenarios such as artificial intelligence cloud, deep learning model training and online reasoning, gene sequencing, weather forecasting, and oil and gas exploration. Inspur also developed management suites such as AIStation, Teye, and ClusterEngine to help users manage and dispatch AI HPC computing resources in a unified manner, realize containerized deployment, visual development and centralized management, and improve resource utilization and work efficiency.
During ISC19, in addition to showcasing product technology and application results in the AI HPC field, Inspur will also work with partners such as Intel to launch the latest AI & HPC converged products and solutions to help users run and manage AI HPC workloads more easily and efficiently, and accelerate R&D and application innovation in AI HPC. Meanwhile, Inspur application experts will share how to use HPC large-scale optimization strategy to improve HPC operating efficiency.
Inspur Booth Event Preview
|16:30-17:00, June 17||Inspur and Intel Jointly Releases New AI HPC Products||
|11:00 am-11:30 am, June 18||Joint speech by Inspur and ThinkParQ|
|11:00 am-11:30 am, June 19||Large-scale optimization strategy for typical HPC workloads|
As one of the important activities of ISC19, the ISC19 Student Cluster Competition Finals was also launched today. 14 college student teams from all over the world will design and build computing clusters with the power consumption limit of 3000W to optimize the specified applications. Inspur sponsored four student teams, namely Taiwan Tsinghua University, Tsinghua University, Sun Yat-sen University and Heidelberg University to participate in the competition.