Performance Evaluation Method for Intelligent Computing Components for Space Applications
Author:
Xie Man12, Wang Lianguo1, Ma Miao1, Zhang Pengfei12
Affiliation:
1. National Space Science Center, Chinese Academy of Sciences, Beijing 100190, China 2. University of Chinese Academy of Sciences, Beijing 100049, China
Abstract
The computational performance requirements of space payloads are constantly increasing, and the redevelopment of space-grade processors requires a significant amount of time and is costly. This study investigates performance evaluation benchmarks for processors designed for various application scenarios. It also constructs benchmark modules and typical space application benchmarks specifically tailored for the space domain. Furthermore, the study systematically evaluates and analyzes the performance of NVIDIA Jetson AGX Xavier platform and Loongson platforms to identify processors that are suitable for space missions. The experimental results of the evaluation demonstrate that Jetson AGX Xavier performs exceptionally well and consumes less power during dense computations. The Loongson platform can achieve 80% of Xavier’s performance in certain parallel optimized computations, surpassing Xavier’s performance at the expense of higher power consumption.
Funder
National Key R&D Program of China
Subject
Electrical and Electronic Engineering,Biochemistry,Instrumentation,Atomic and Molecular Physics, and Optics,Analytical Chemistry
Reference29 articles.
1. Lovelly, T.M., Bryan, D., Cheng, K., Kreynin, R., George, A.D., Gordon-Ross, A., and Mounce, G. (2014, January 1–8). A framework to analyze processor architectures for next-generation on-board space computing. Proceedings of the 2014 IEEE Aerospace Conference, Big Sky, MT, USA. 2. Doyle, R., Goforth, M., Some, R., Powell, W., Whitaker, W., Bergman, L., Johnson, M., and Lowry, M. (2013). Human and Robotic Space Mission Use Cases for High-Performance Spaceflight Computing, AIAA. 3. Doyle, R., Some, R., Powell, W., Mounce, G., Goforth, M., Horan, S., and Lowry, M. (2014, January 17–19). High Performance Spaceflight Computing; Next-Generation SpaceProcessor: A Joint Investment of NASA and AFRL. Proceedings of the International Symposium on Artificial Intelligence, Robotics, and Automation in Space (i-SAIRAS), Montreal, QC, Canada. 4. Mounce, G., Lyke, J., Horan, S., Powell, W., Doyle, R., and Some, R. (2016, January 5–12). Chiplet based approach for heterogeneous processing and packaging architectures. Proceedings of the 2016 IEEE Aerospace Conference, Big Sky, MT, USA. 5. Can, L. (2019). Research on GPU Array Parallel Algorithm for Geometric Correction of Remote Sensing Images. [Master’s Thesis, University of Electronic Science and Technology of China].
|
|