Floating Point Lab Report

CHAPTER 1
INTRODUCTION
In the design of digital processer and application specific system Digital arithmetic operations are very important. Floating Point Arithmetic and logic Unit in Digital System FPGA implementation has been the most widely researched area in recent years separately and together. Separately many new architectures of ALU implementation have been proposed. As far as floating point unit or DSP block is concerned, it has been one of the most complex units for implementation on FPGA. But with the help of VHDL these implementations have become very easy now a day. The computation has become very easy for very long numbers Floating point describes a system for

Floating point describes a system for representing numbers that would be too large or too small can be represented as integers. Floating point representation is able to retain its resolution and accuracy compared to fixed point representation. IEEE specified standard for floating-point representation known as IEEE 754 and given in late 1985.
Fixed point numbers go through the limited range and accuracy. For a given word length fixed point and floating point both representations give equal distinct numbers for that given world length. The difference between both the fixed point numbers representation and floating point numbers representation is that is in fixed point representation the spacing between the numbers is equal, so smaller numbers when truncated or rounded give a much larger error than the larger numbers. However the floating point numbers representation gives different spacing between numbers. We get denser distances between numbers when the number is small and sparser distance for larger numbers. So the absolute representation error increases with larger