The design and implementation of a very fast experimental pipelining computer
The high speed potential of I.C. components can be exploited by shortening the pipeline clock period. Although there are some factors which dominate the shortening, the design of an experimental computer employs the principle of maximum time difference at the system level to determine the clock period and the integrated consideration of architecture, logic design and engineering layout to achieve a system clock period of 9.8 ns using conventional ECL chips of 2ns gate delay. The multiplier in this model, which is constructed with 0.7 ns gate delay chips, can work at a clock period of 5.5 ns.
KeywordsPropagation Delay Clock Period Logic Design Gate Delay Single Board Computer
Unable to display preview. Download preview PDF.
- K. Hwang, Computer Arithmetic Principles, Architecture and Design, John Wiley & Sons Inc., 1979.Google Scholar
- L.W. Cotten, Maximum-rate Pipeline System, AFIPS Proc. SJCC, 1969, 581–586.Google Scholar
- P.M. Kogge, The Architecture of Pipelined Computers, McGrow-Hill, 1981.Google Scholar
- Makoto Tatsuriet al., An ECL 5000-gate array with 190-ps gate delay,Solid-State Circuits, SC-21:2 (1986).Google Scholar
- FlOOK ECL Data Book, Fairchild Camera and Instrument Corporation, 1982.Google Scholar