At NVIDIA GTC FALL 2021, the ultimate AI conference from November 8th-14th, Inspur Information presents the latest developments in AI and Edge Computing.
The Most Advanced GPU Platform Empowers AI Model Innovation
Lei Wang /Senior Product Manager
Learn about Inspur's most advanced GPU platform, including their industry application.
The Development Trend and Industry Applications of Intelligent Edge Computing
Kris Liu /Senior Product Manager
Learn about Inspur's edge server innovation, including their industry application.
Analysis of MLPerf Competition Resnet50 Training with Inspur GPU server
Baihong Li /Senior Architect
Discover what is behind the Inspur GPU servers that broke multiple records at the latest MLPerf training benchmark tests.
Yuan 1.0:Large-scale pre-trained language model in Zero-Shot and Few-Shot Learning
Shawn Wu /Principle researcher, Inspur AI Research Institute
Get introduced to Inspur’s Yuan 1.0, the world’s largest Chinese language model with 245.7 billion parameters and 5 TB of datasets.
As world’s leading AI computing provider, Inspur offers a broad range of cutting-edge computing platforms to empower some of the most challenging AI supercomputing tasks the world facing today.
Leading configurations, topology switching with one key
The NF5468M6 can be loaded with 2 Intel 3rd generation Xeon SP processors and up to 8 A100/A30 or 16 A10 GPUs in 4U space, which is widely used in AI public cloud and enterprise AI cloud platform, video AI and other scenarios, responding flexibly to the performance tuning needs of different AI application scenarios.Learn More
Excellent performance output, excellent eco-adaptation
The NF5468A5 supports the most current AI computing technologies and provides superior performance output, continuing Inspur's high-quality and highly compatible design with a mature ecosystem that provides professional AI customers around the world with a powerful and reliable infrastructure support.Learn More
Excellent versatility, exceptional reliability
NF5280M6 is a 2U dual-socket high-end flagship rackmount server, featured with robust computing performance and ultimate compatibility and scalability. Meeting the configuration requirements of various industries, it is suitable for data analysis and processing, distributed storage in deep learning training, etc.Learn More
4U 2-Socket Nvlink Server
6U 2-Socket Nvlink Server
2U-4Node liquid cooled server optimized for high density data center and HPC
i24LM6 is suitable for a wide variety of compute-intensive workloads including HPC, high-performance data analytics and more.
Extreme Density FPGA Accelerator Card
Built on Intel® Arria® 10 FPGA internals, the Inspur F10A is designed to accelerate a range of applications, including big data analytics, Artificial Intelligence, video transcoding, high performance computing, genomics and more. With breakthrough performance and power efficiency, it delivers real-time insight from cloud to the edge.
High-Density VPU Accelerator
M10A is Inspur's new generation high-density VPU(Video Processing Unit) accelerator for full-scenario video AI processing applications, which supports 16ch 1080P@30fps real time video codec with 8W ultra-low power consumption, and supports the powerful H.265 video encoding algorithms, which doubles the network bandwidth utilization, and reduces the CPU workload as low as 2%. M10A can fully support the leading multimedia framework FFMPEG realizing almost zero migration cost from current system.
Based on its diverse experiences in AI and HPC, Inspur offers a series of agile management tools, powering AI HPC applications from development to production.
Agile AI Development Platform
Inspur AIStation is designed to provide complete AI development software stack, unified management of AI computing resources and simplified shift of AI models from development to production. It has been adopted by a wide range of industries users, accelerating AI transformation.
Efficient HPC Cluster Management Tool
Inspur ClusterEngine offers integrated management of HPC clusters, including hardware monitoring, job scheduling, HPC and Bid Data applications management and so on. It has been widely adopted to improve resource utilization of the entire HPC system within one overview dashboard.