This Article 
 Bibliographic References 
 Add to: 
Buffered Banks in Multiprocessor Systems
April 1995 (vol. 44 no. 4)
pp. 518-530

Abstract—A memory design based on logical banks is analyzed for shared memory multiprocessor systems. In this design, each physical bank is replaced by a logical bank consisting of a fast register and subbanks of slower memory. The subbanks are buffered by input and output queues which substantially reduce the effective cycle time when the reference rate is below saturation. The principal contribution of this work is the development of a simple analytical model which leads to scaling relationships among the efficiency, the bank cycle time, the number of processors, the size of the buffers, and the granularity of the banks. These scaling relationships imply that if the interconnection network has sufficient bandwidth to support efficient access using high-speed memory, then lower-speed memory can be substituted with little additional interconnection cost. The scaling relationships are shown to hold for a full datapath vector simulation based on the Cray Y-MP architecture. The model is used to develop design criteria for a system which supports 192 independent reference streams, and the performance of this system is evaluated by simulation over a range of loading conditions.

[1] D.H. Bailey,“Vector computer memory bank contention,” IEEE Trans. Computers, vol. 36, pp. 293-298, 1987.
[2] F.A. Briggs,“Effects of buffered memory requests in multiprocessorsystems,” Proc. ACM-SIGMETRICS Conf. Simulation, Measurements, and Modelingof Computer Systems, pp. 434-442, 1979.
[3] F.A. Briggs and E.S. Davidson,“Organization of semiconductor memories forparallel-pipelined processors,” IEEE Trans. Computers, vol. 26, pp. 162-169, 1977.
[4] D. Bondurant,“Enhanced dynamic RAM,” IEEE Spectrum, vol. 29, no. 10, pp. 49, 1992.
[5] T. Cheung and J.E. Smith,“A simulation study of the Cray X-MP memorysystem,” IEEE Trans. Computers, vol. 35, pp. 613-622, 1986.
[6] M.P. Farmwald and D. Mooring,“A fast path to one memory,” IEEE Spectrum, vol. 29, no. 10, pp. 50-51, 1992.
[7] D.T. Harper and J.R. Jump,“Vector access performance in parallel memoriesusing a skewed storage scheme,” IEEE Trans. Computers, vol. 36, pp. 1440-1449, 1987.
[8] D.T. Harper,“Address transformations to increase memory performance,” Proc. 1989 Int’l Conf. Parallel Processing, pp. 237-241, 1989.
[9] C.A. Hart,“Dynamic RAM as secondary cache,” IEEE Spectrum, vol. 29, no. 10, pp. 48, 1992.
[10] R. Jain,The Art of Computer Systems Performance Analysis, John Wiley&Sons, 1991.
[11] N. Kushiyama,Y. Watanabe,T. Ohsawa,K. Muraoka,Y. Nagahama,, and T. Furuyama,“A 12-MHz data cycle 4-Mb DRAM with pipeline operation,” IEEE J. Solid-State Circuits, vol. 26, no. 4, pp. 479-482, 1991.
[12] D.H. Lawrie and C.R. Vora,“The prime memory system for array access,” IEEE Trans. Computers, vol. 31, pp. 435-442, 1982.
[13] D. Lee,“Scrambled storage for parallel memory systems,” Proc. 15th Int’l Conf. Computer Architecture, pp. 232-239, 1988.
[14] B. Prince,R. Norwood,J. Hartigan,, and W.C. Vogley,“Synchronous dynamic RAM,” IEEE Spectrum, vol. 29, no. 10, pp. 44-48, 1992.
[15] K.A. Robbins and S. Robbins,The Cray X-MP/Model 24: A Case Study inPipelined Architecture and Vector Processing,Berlin Heidelberg: Springer Lecture Notes in Computer Science, vol. 374, 1989.
[16] K.A. Robbins and S. Robbins,“Bus conflicts for logical memory banks on aCray Y-MP type processor system,” 1991 Int’l Conf. Parallel Processing, pp. 21-24, 1991.
[17] K.A. Robbins and S. Robbins,“Dynamic behavior of memory reference streams for the Perfect Club benchmarks,” Proc. 1992 Int’l Conf. Parallel Processing, pp. 48-52, 1992.
[18] K.A. Robbins and S. Robbins,“Relationships between average and real memory behavior,” J. Supercomputing, vol. 8, pp. 209-232, 1994.
[19] A. Seznec and Y. Jegou,“Optimizing memory throughput in a tightly coupled multiprocessor,” Proc. 1987 Int’l Conf. Parallel Processing, pp. 344-346, 1987.
[20] J.E. Smith and W.R. Taylor,“Accurate modeling of interconnection networks in vector supercomputers,” 1991 Int’l Conf. Supercomputing, pp. 264-273, 1991.
[21] G.S. Sohi,“High-bandwidth interleaved memories for vector processors—Asimulation study,” IEEE Trans. Computer Systems, vol. 42, pp. 34-44, 1993.
[22] C.B. Stunkel,B. Janssens,, and W.K. Fuchs,“Address tracing for parallel machines,” Computer, pp. 31-38, 1991.

Index Terms:
Buffered memories, logical memory banks, memory conflicts, vector processors, Cray Y-MP.
Steven Robbins, Kay A. Robbins, "Buffered Banks in Multiprocessor Systems," IEEE Transactions on Computers, vol. 44, no. 4, pp. 518-530, April 1995, doi:10.1109/12.376167
Usage of this product signifies your acceptance of the Terms of Use.