ABSTRACT
Assessing the performance and software interactions of emerging parallel input/output systems is a critical first step in input/output software tuning. Moreover, understanding the system response to well-understood, synthetic input/output patterns is itself a prelude to analysis of more complex application input/output patterns. We have conducted a series of experiments to measure the performance of the CM-5's new Scalable Disk Array (SDA) and Scalable Parallel File System (SFS) using the file system interfaces provided by the data parallel CM Fortran and message passing CMMD programming models. The results of these experiments suggest that the CM-5's parallel input/output systems is an improvement over its predecessor, the CM-2 Data Vault. However, network bandwidth can be a bottleneck for the data reordering phase of input/output operations.
- 1.BATCHER, K. E. Deign of a Massively Parallel Processor. IEEE Transactions on Computers C-29, (Sept. 1980), 836-842.Google ScholarDigital Library
- 2.BEST, M. Personal Communication.Google Scholar
- 3.BORDAWEKAR, R., CHOUDHARY, A., AND DEL ROSARIO, J. M. An Experimental Performance Evaluation of Touchstone Delta Concurrent File System. In Proceedings of the International Conference on Supercomputing (July 1993), pp. 367-377. Google ScholarDigital Library
- 4.BORDAWEKAR, R., DEL ROSARIO, J. M., AND CHOUDHARY, A. Design and Evaluation of Primitives for Parallel I/O. In Supercomputing 1993 (Nov 1993), pp. 452-461. Google ScholarDigital Library
- 5.FINBBERG, S. A. Implementing the NHT- 1 Application I/O Benchmark. In Workshop on I/O in Parallel Computer Systems (April 1993).Google ScholarDigital Library
- 6.HPFF. High-Peformance Fortran Language Specification, version 1.0. Tech. rep., High Performance Fortran Forum, May 1993.Google Scholar
- 7.INTEL SUPERCOMPUTER SYSTEMS DIVI- SION. Paragon User's Guide, Oct 1993.Google Scholar
- 8.KWAN, T. T. Performance Evaluation of the CM-5. Master's thesis, University of Illinois at Urbana-Champaign, Department of Computer Science, April 1994.Google Scholar
- 9.KWAN, T. T., TOTTY, B. K., AND REED, D.A. Communication and Computation Performance of the CM-5. In Supercomputing 1993 (Nov 1993), pp. 192-201. Google ScholarDigital Library
- 10.LEISERSON, C. E., ET AL. The Network Architecture of the Connection Machine CM- 5. In Proceedings of Parallel Algorithms and Architectures Symposium (May 1992), pp. 272-285. Google ScholarDigital Library
- 11.LoVERSO, S. J., ISMAN, M., NANOPOU- LOS, A., NESHEIM, W., MILNE, E. D., AND WHEELER, R. sfs: A Parallel File System for the CM-5. In Proceedings of the 1993 Summer Usenix Conference (1993), pp. 291-305.Google Scholar
- 12.REEDD, D. A., AYDT, R. A., NOE, R. J., ROTH, P. C., SHIELDS, K. A., SCHWARTZ, B. W., AND TAVERA, L. F. Scalable Performance Analysis: The Pablo Performance Analysis Environment. In Proceedings of the Scalable Parallel Libraries Conference (1993), A. Skjellum, Ed., IEEE Computer Society.Google Scholar
- 13.THINKING MA- CHINES CORPORATION. CMMD Reference Manual Version, 3.0 Beta, Dec 1992.Google Scholar
- 14.THINKING MACHINES CORPORATION. CM- 5 I/O System Programming Guide, Version 7.2, September 1993.Google Scholar
- 15.THINKING MACHINES CORPORATION. CM Fortran Utility Library Reference Manual, Version 2.0 Beta, January 1993.Google Scholar
Index Terms
- Performance of the CM-5 scalable file system
Recommendations
Performance comparison of the CM-5 and Intel Touchstone Delta for data parallel operations
SPDP '93: Proceedings of the 1993 5th IEEE Symposium on Parallel and Distributed ProcessingData parallel programs on MIMD machines are often structured as alternating phases of local computation and global communication such as reduction, synchronization, broadcast, etc. This paper describes the performance comparison of data parallel ...
Implementation and performance of a parallel file system for high performance distributed applications
HPDC '96: Proceedings of the 5th IEEE International Symposium on High Performance Distributed ComputingDedicated cluster parallel computers (DCPCs) are emerging as low-cost high performance environments for many important applications in science and engineering. A significant class of applications that perform well on a DCPC are coarse-grain applications ...
Scalable performance of the Panasas parallel file system
FAST'08: Proceedings of the 6th USENIX Conference on File and Storage TechnologiesThe Panasas file system uses parallel and redundant access to object storage devices (OSDs), per-file RAID, distributed metadata management, consistent client caching, file locking services, and internal cluster management to provide a scalable, fault ...
Comments