A bipolar population counter using wave pipelining to achieve 2.5* normal clock frequency

A bipolar LSI chip which achieves 2.5 times the normal clock frequency by means of wave pipelining without the use of additional storage elements is described. In wave pipelining, multiple coherent waves of data are placed between storage elements by clocking the circuit faster than the propagation delay of the combinational logic. If all the propagation paths from the combinational circuit inputs to outputs have approximately the same delay, each wave propagates uniformly to the outputs without interfering with adjacent waves. The wave pipelining concept has been tested using a demonstration chip. Compared to an implementation using ordinary pipelining, a wave-pipelined circuit reduces the latency, area, and clock distribution required by pipeline registers or latches. >

[1]  Giovanni De Micheli,et al.  Designing high-performance digital circuits using wave pipelining: algorithms and practical experiences , 1993, IEEE Trans. Comput. Aided Des. Integr. Circuits Syst..

[2]  L. W. Cotten Maximum-rate pipeline systems , 1969, AFIPS '69 (Spring).

[3]  Giovanni De Micheli,et al.  Inserting active delay elements to achieve wave pipelining , 1989, 1989 IEEE International Conference on Computer-Aided Design. Digest of Technical Papers.

[4]  S. F. Anderson,et al.  The IBM system/360 model 91: floating-point execution unit , 1967 .

[5]  Maciej J. Ciesielski,et al.  Placement for clock period minimization with multiple wave propagation , 1991, 28th ACM/IEEE Design Automation Conference.

[6]  C. A. Zukowski,et al.  A matched-delay CMOS TDM multiplexer cell , 1988, Proceedings 1988 IEEE International Conference on Computer Design: VLSI.

[7]  R. Lane,et al.  Single polysilicon layer advanced super high-speed BiCMOS technology , 1989, Proceedings of the Bipolar Circuits and Technology Meeting.

[8]  Wentai Liu,et al.  Theoretical and Practical Issues in CMOS Wave Pipelining , 1991, VLSI.