SSE Stock Code：603019
You can also follow the Sugon WeChat official account
Scan with WeChat to log in. Access materials more easily.
January 2025
Service Hotline：400-810-0466
Live Support, Online Repair, Warranty Inquiry, Device Binding, Complaints
GPU platforms that turn internet-scale data into deep-learning capacity.
Sugon Xmachine is designed for deep-learning users — high parallelism, high throughput, and low latency for strong compute across scenarios.
ParaStor distributed storage supplies aggregate I/O bandwidth, with a reliable system that scales linearly.
A containerized learning platform for fast environment deployment and job assignment, cutting the complexity of installing and migrating many applications.
Covers AI and HPC, with rich scheduling policies and flexible resources that raise cluster utilization for more efficient, elastic training.
The compute system of Sugon’s GPU deep-learning platform uses Sugon’s new-generation XMachine high-performance GPU servers to form large-scale GPU trainingand inference clusters, outputting powerful computing power;Compute data, logs, and model data are stored uniformly in Sugon ParaStor300 distributed parallel storage, enabling unified global-fileaccess and concurrent read/write;The system supports Caffe/TensorFlow and other mainstream deep-learning frameworksand uses container technologies, providing dataset management, model management, trainingand other services. It helps users with multi-group resource allocation, rapid development-environment setup, and flexible application migration. It lets users on the clustereasily deploy deep-learning applications, track experiments and training, and publish models,whilewithout having to care about tedious deployment and O&M, focusing on core business。
Finer-grained control and on-demand GPU allocation that raises utilization and delivers compute efficiently and economically.
One-click environment deployment and graphical management that simplify building a deep-learning platform for internet companies.
Optimization across distributed systems, parallel machine-learning execution, and algorithm toolkits to help tune applications.
+86 400-810-0466
international@sugon.com
津公网安备 12011602000521号
+86 400-810-0466
international@sugon.com
津公网安备 12011602000521号
Register /
