All modern processor, including microprocessor, DSP contains ALU.
The computing efficiency of ALU an adder is the basic building block for an ALU which performs arithmetic as well as logic operations.
The existing adders cannot meet to the expected optimization so we
proposed 4 types of parallel prefix adders that uses prefix operation in order to do efficient addition. Introduction
Binary addition is most important operation in computer
arithmetic.
VLSI integer adders are critical elements in general purpose micro
processor and digital signal processor since they are employed in ALU’s in floating point arithmetic datapaths and in address generation units. Parallel prefix adder
Parallel prefix adder perform parallel
addition. PPA carries out three necessary and important steps. 1. Pre processing 2. Carry graph 3. Post processing Types of parallel prefix adder
There are four types of parallel prefix adders they are as follows
Sklansky adder
Kogge stone adder
Brent kung adder
Ladner Fischer adder
Sklansky adder • Sklansky adder is also called as divide and conquer tree. • The fan out of this adder increases significantly from the inputs to outputs along the critical path which accounts for large amount of latency. • This degrades the performance of the structure when the number of bits of the adder become larger. Kogge stone adder • It has both optimal depth and low fan out but produces massively complex circuit realization and also accounts for large no of interconnect. • Adders implemented using this technique possess regular layout and a controlled fan out of two. Brent kung adder
It uses less computation nodes but possess
maximal depth which accounts for increased latency. An inverse carry tree is added to compute intermediate carries. Its wire complexity is much less than that of a kogge stone adder. Ladner fischer adder • It is the family of network between Sklansky adder and Brent Kung adder which computes prefixes for the odd numbered bits and again uses one more stage to ripple into the even positions. • This method reduces the maximum fan out for computation nodes in the critical path. Conclusion
The parallel prefix adders are faster because of less delay
and area efficient compared to other basic adders.