Javascript must be enabled to continue!
Design and Performance Evaluation of HEPS Data Center Network
View through CrossRef
Among the 15 beamlines in the first phase of the High-Energy Photon Source (HEPS) in China, the maximum peak data generation volume can reach 1 PB per day, with the maximum peak data generation rate reaching 3.2 Tb/s. This poses significant challenges to the underlying network system. This paper designed a high-performance and scalable network architecture based on RoCE (RDMA over Converged Ethernet) to meet the needs of storage, computing, and analysis of HEPS scientific data. Test results show that the RoCE-based HEPS data center network system achieves high bandwidth and low latency characteristics, stably maintains reliable transmission performance during the interaction of scientific data storage, computing, and analysis, while also exhibiting good scalability and being able to adapt to the future expansion needs of HEPS beamlines.
Title: Design and Performance Evaluation of HEPS Data Center Network
Description:
Among the 15 beamlines in the first phase of the High-Energy Photon Source (HEPS) in China, the maximum peak data generation volume can reach 1 PB per day, with the maximum peak data generation rate reaching 3.
2 Tb/s.
This poses significant challenges to the underlying network system.
This paper designed a high-performance and scalable network architecture based on RoCE (RDMA over Converged Ethernet) to meet the needs of storage, computing, and analysis of HEPS scientific data.
Test results show that the RoCE-based HEPS data center network system achieves high bandwidth and low latency characteristics, stably maintains reliable transmission performance during the interaction of scientific data storage, computing, and analysis, while also exhibiting good scalability and being able to adapt to the future expansion needs of HEPS beamlines.
Related Results
Non-Recommended Publishing Lists: Strategies for Detecting Deceitful Journals
Non-Recommended Publishing Lists: Strategies for Detecting Deceitful Journals
Abstract
The rapid growth of open access publishing (OAP) has significantly improved the accessibility and dissemination of scientific knowledge. However, this expansion has also c...
Performative Microforests
Performative Microforests
The design of office buildings can substantially improve the building, social, and ecological performance of office building projects. However, existing research on improving the p...
Network Automation
Network Automation
Purpose: The article "Network Automation in the Contemporary Economy" explores the concepts and methods of effective network management. The application stack, Jinja template engin...
Innovations in Multicore Network Processor Design for Enhanced Performance
Innovations in Multicore Network Processor Design for Enhanced Performance
The rapid expansion of network traffic, driven by the proliferation of internet-connected devices and the growing demand for high-speed data transmission, has intensified the need ...
High Energy Particle Spectrometer for ESA Lagrange mission
High Energy Particle Spectrometer for ESA Lagrange mission
<p>The LGR High Energy Particle Spectrometer HEPS for the ESA Lagrange mission belongs to the satellite in-situ instrument suite. The satellite will be placed at the ...
The synergistic effect of ego-network stability and whole network position: a perspective of transnational coopetition network
The synergistic effect of ego-network stability and whole network position: a perspective of transnational coopetition network
PurposeThe authors selected global automobile manufacturing firms whose sales ranked within 100 in the five years from 2014 to 2018 in the Factiva database to examine how the chara...
"Best Tradition": CREATE, JCSEE and the Program Evaluation Standards
"Best Tradition": CREATE, JCSEE and the Program Evaluation Standards
Background: Evaluation “is a task in the best tradition of the most abstract theoretical science as well as the most practical applied science” (Scriven, 1968, p .9). The Program E...

