Abstract.
The data generated by a High Performance Computing(HPC) server is very large. For example, a combustion chemical reaction flow simulation generates hundreds gigabytes data. In order to analyze this enormous amount of data, a huge HPC server and huge pre/post systems are both necessary. In this paper, we have benchmarked the I/O performance of two applications on NS-III. One application environment is the GSN library visualization system and the other is FORTRAN I/O library on NS-III. NS-III is the name of a supercomputer system located at National Aerospace Laboratory. That system has a nine-teraflops compute system, a huge visualization system, and a high-speed mass storage system. These servers are connected with GSN, and/or crossbar-network. In addition to benchmark results, we also demonstrate the user-programming interface of the GSN library that realizes a file sharing.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Fujita, N., Matsuo, Y.: High-Speed Mass Storage System of Numerical Simulator III and its Basic I/O Performance Benchmark. Parallel Computational Fluid Dynamics, pp. 157–164. Elsevier Science B.V, Amsterdam (2003)
Fujita, N., Ookawa, H.: Storage Devices, Local File System and Crossbar Network File System Characteristics, and 1 Terabyte File IO Benchmark on the “Numerical Simulator III”. In: Proc. of 20th IEEE Symposium on Mass Storage Systems and Technologies/11th NASA Goddard Conference on Mass Storage Systems and Technologies, pp. 72-76 (2003)
Jones, T., Sarnowska, B., Lovato, F., Magee, D., Koth, J.: Architecture, Implementation, and Deployment of a High Performance, High Capacity Resilient Mass Storage Server(RMSS). In: Proc. of 18th IEEE Symposium on Mass Storage Systems and Technologies/9th NASA Goddard Conference on Mass Storage Systems and Technologies, pp. 41–65 (2001)
Stone, N.T.B., Scott, J.R., et al.: Mass Storage on the Terascale Computing System. In: Proc. of 18th IEEE Symposium on Mass Storage Systems and Technologies/9th NASA Goddard Conference on Mass Storage Systems and Technologies, pp. 67– 77 (2001)
Andrews, P., Sherwin, T., Hazlewood, V.: High-Speed Data Transfer via HPSS using Striped Gigabit Ethernet Communications. In: Proc. of 18th IEEE Symposium on Mass Storage Systems and Technologies/9th NASA Goddard Conference on Mass Storage Systems and Technologies, pp. 127–133 (2001)
HiPPI-6400/Gigabyte System Network, http://public.lanl.gov/radiant/research/architecture/hippi.html
High Performance Networking Forum: http://www.hnf.org/
If you are interested in these functions, please refer “CeNSS User Reference Guide” on https://censs.nal.go.jp/
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Fujita, N., Ookawa, H. (2003). The GSN Library and FORTRAN Level I/O Benchmarks on the NS-III HPC System. In: Veidenbaum, A., Joe, K., Amano, H., Aiso, H. (eds) High Performance Computing. ISHPC 2003. Lecture Notes in Computer Science, vol 2858. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-39707-6_41
Download citation
DOI: https://doi.org/10.1007/978-3-540-39707-6_41
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-20359-9
Online ISBN: 978-3-540-39707-6
eBook Packages: Springer Book Archive