This Article 
   
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
Scalar Memory References in Pipelined Multiprocessors: A Performance Study
January 1992 (vol. 18 no. 1)
pp. 78-86

Interleaved memories are essential in pipelined computers to attain high memory bandwidth. As a memory bank is accessed, a reservation is placed on the bank for the duration of the memory cycle, which is often considerably longer than the processor cycle time. This additional parameter, namely, the bank reservation time or the bank busy time, adds to the complexity of the memory model. For Markov models, exact solutions are not feasible even without this additional parameter due to the very large state space of the Markov chain. The authors develop a Markov model which explicitly tracks the bank reservation time. Because only one processor and the requested bank are modeled, the transition probabilities are not known and have to be approximated. The performance predicted by the model is in close agreement with simulation results.

[1] D. H. Bailey, "Vector computer memory bank contention,"IEEE Trans. Computers, vol. C-36, pp. 293-298, Mar. 1987.
[2] J. M. Frailong, W. Jalby, and J. Lenfant, "XOR-schemes: a flexible data organization in parallel memories," inProc. 1985 Int. Conf. Parallel Process., Aug. 1985, pp. 276-283.
[3] A. Norton and E. Melton, "A class of boolean linear transformations for conflict-free power-of-two access," inProc. 1987 Int. Conf. Parallel Process., Aug. 1987, pp. 247-254.
[4] M. Balakrishnan, R. Jain, and C. S. Raghavendra, "On array storage for conflict-free memory access for parallel processors," inProc. 1988 Int. Conf. Parallel Process., Aug. 1988, pp. 103-107.
[5] D. Lee, "Scrambled storage for parallel memory systems," inProc. 15th Ann. Int. Symp. Computer Arch., May 1988.
[6] G. S. Sohi, "High-bandwidth interleaved memories for vector processors--a simulation study," Computer Sci. Dept., Univ. Wisconsin-Madison, Tech. Rep. 790, Sept. 1988.
[7] B. R. Rau, M. S. Schlansker, and D. W. L. Yen, "The Cydra 5 stride-insensitive memory system," inProc. 1989 Int. Conf. Parallel Process., Aug. 1989, pp. 242-246.
[8] R. Raghavan and J. P. Hayes, "On randomly interleaved memories," inProc. Supercomputing '90 Conf.(New York), Nov. 1990.
[9] T. Cheung and J. E. Smith, "A simulation study of the CRAY X-MP memory system,"IEEE Trans. Computers, vol. C-35, pp. 613-622, July 1986.
[10] W. Oed and O. Lange, "On the effective bandwidth of interleaved memories in vector processing systems,"IEEE Trans. Comput., vol. C-34, no. 10, pp. 949-957, Oct. 1985.
[11] H. Hellerman,Digital Computer Systems, 2nd ed. New York: McGraw-Hill, 1973.
[12] D. E. Knuth and G. S. Rao, "Activity in interleaved memory,"IEEE Trans. Computers, vol. C-24, pp. 943-944, Sept. 1975.
[13] C. V. Ravi, "On the bandwidth and interference in interleaved memory systems,"IEEE Trans. Computers, vol. C-21, pp. 899-901, Aug. 1972.
[14] D. Y. Chang, D. J. Kuck, and D. H. Lawrie, "On the effective bandwidth of parallel memories,"IEEE Trans. Computers, vol. C-26, pp. 480-489, May 1977.
[15] D. P. Bhandarkar, "Analysis of memory interference in multiprocessors,"IEEE Trans. Computers, vol. C-24, pp. 897-908, Sept. 1975.
[16] F. S. Baskett and A. J. Smith, "Interference in multiprocessor computer systems with interleaved memory,"Commun. ACM, vol. 19, pp. 327-334, Jun. 1976.
[17] G. Chiola, M. A. Marsan, and G. Balbo, "Product-form solutions techniques for the performance analysis of multiple-bus multiprocessor systems with nonuniform memory references,"IEEE Trans. Computers, vol. C-37, pp. 532-540, May 1988.
[18] M. A. Holliday and M. K. Vernon, "Exact performance estimates for muitiprocessor memory and bus interference,"IEEE Trans. Comput., vol. C-36, pp. 76-85, Jan. 1987.
[19] C. H. Hoogendoorn, "A general model for memory interferences in multiprocessors,"IEEE Trans. Computers, vol. C-26, pp. 998-1005, Oct. 1977.
[20] K. B. Irani and I. H. Onyuksel, "A closed-form solution for the performance analysis of multiple-bus multiprocessor systems,"IEEE Trans. Computers, vol. C-33, pp. 1004-1012, Nov. 1984.
[21] J. H. Patel, "Processor-memory interconnections for multiprocessors,"IEEE Trans. Computers, vol. C-30, pp. 771-780, Oct. 1981.
[22] B. R. Rau, "Interleaved memory bandwidth in a model of multiprocessors,"IEEE Trans. Computers, vol. C-28, pp. 678-681, Sept. 1979.
[23] A. S. Sethi and N. Deo, "Interference in multiprocessor memory systems with localized memory access probabilities,"IEEE Trans. Computers, vol. C-28, pp. 157-163, Feb. 1979.
[24] C. E. Skinner and J. R. Asher, "Effects of storage contention on system performance,"IBM Syst. J., vol. 8, pp. 319-333, 1969.
[25] D. Towsley, "Approximate models of multiple bus multiprocessor systems,"IEEE Trans. Computers, vol. C-35, pp. 220-228, Mar. 1986.
[26] D. W. L. Yen, J. H. Patel, and E. S. Davidson, "Memory interference in synchronous multiprocessor systems,"IEEE Trans. Computers, vol. C-31, pp. 1116-1121, Nov. 1982.
[27] T. N. Mudge and H. B. Al-Sadoun, "Memory interference models with variable connection time,"IEEE Trans. Computers, vol. C-33, pp. 1033-1038, Nov. 1984.
[28] F. A. Briggs and E. S. Davidson, "Organization of semiconductor memories for parallel-pipelined processors,"IEEE Trans. Computers, vol. C- 26, pp. 162-169, Feb. 1977.
[29] I. Y. Bucher and D. A. Calahan, "Access conflicts in multiprocessor memories queueing models and simulation studies," inProc. 1990 Int. Conf. Supercomput.(Amsterdam, The Netherlands), June 1990, pp. 428-438.

Index Terms:
scalar memory references; pipelined multiprocessors; pipelined computers; high memory bandwidth; memory bank; memory cycle; processor cycle time; bank reservation time; bank busy time; Markov models; state space; Markov chain; transition probabilities; simulation results; Markov processes; parallel machines; performance evaluation; pipeline processing; probability; storage management
Citation:
R. Ganesan, S. Weiss, "Scalar Memory References in Pipelined Multiprocessors: A Performance Study," IEEE Transactions on Software Engineering, vol. 18, no. 1, pp. 78-86, Jan. 1992, doi:10.1109/32.120318
Usage of this product signifies your acceptance of the Terms of Use.