Rate and give feedback:
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document.
Note: Even the most advanced machine translation cannot match the quality of professional translators.
Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).
Scenario
An HPC system is built to improve the computing speed to a manner of tera operations per second (TOPS), and is mainly used for large-scale scientific data computing and processing of huge volumes of data, for example, scientific research, weather forecasting, computational simulation, military research, biopharmacy research, gene sequencing, and image processing. The HPC scenario has the following characteristics:
- A single task is split into multiple subtasks for parallel processing to improve computing efficiency. As the computing performance of servers continuously improves, more pressure is placed on the network performance and the waiting time for communication between compute nodes accounts for a large proportion of the total task duration. In the HPC scenario, the intelligent lossless network needs to guarantee low latency to reduce the communication waiting time.
- HPC applications use the N:N traffic model during parallel computing (servers use similar working modes) and the N:1 incast traffic model during task result synchronization.
- In most HPC scenarios, CPU chips are used for computing, and a large number of servers are deployed. Generally, servers are single-homed to leaf switches through 100GE links to reduce the latency.