Successively higher levels of device integration in microelectronics have caused reduction of power dissipation to become a primary design goal. In many integrated circuit designs, the multiplier is a critically important block, due to its high latency as well as high power characteristics. In this work, we investigate sources of power dissipation in various implementations of small-bit-width digital multipliers. Techniques such as gate-level architectural designs and latch insertion for minimization of glitching are examined. We assess the utility of several techniques for power reduction, and introduce the application of bit polarity inversion to minimize circuit count. The impact of physical design on multiplier power characteristics is presented, along with an exposition of relevant power and delay trade-offs.
Recommendations
Design and implementation of low-power digit-serial multipliers
ICCD '97: Proceedings of the 1997 International Conference on Computer Design (ICCD '97)Digit-serial architectures obtained using traditional unfolding techniques cannot be pipelined beyond a certain level because of the presence of feedback loops. In this paper, a novel design methodology is presented which permits bit-level pipelining of ...
Low power multipliers based on new hybrid full adders
Five hybrid full adder designs are proposed for low power parallel multipliers. The new adders allow NAND gates to generate most of the multiplier partial product bits instead of AND gates, thereby lowering the power consumption and the total number of ...
Comparative analysis of various types of multipliers for effective low power
AbstractNowadays, low power VLSI multiplier with high frequencies shows an important role in the VLSI field. Analysis and comparison of different types of multipliers such as wallace tree, array and baugh wooley multiplier were conceded out in ...
Graphical abstractDisplay Omitted
Highlights- Analysis of different types of multipliers was conceded out in this design.
- To ...