Introduction to Fixed Point representation

Tutorials on Fixed Point representation & arithmetic:

Index

1. What is fixed point number?
2. Why fixed point representations.
3. Fixed point representation, Q-formats.
4. Finally, most important one

• How to select Q-format for a given floating code.
• How to convert floating point to fixed point numbers
  
• Example source code, explaining floating to fixed point conversion’s
• Notes on over flow, underflow and saturation conditions.

What is Fixed-Point?
In simple words, representing real number using Integer data type and hardware. Fixed-point is an alternative to float-point, a more well known method in Embedded and Digital Signal Processing filed (Audio, Video, Image codecs etc).

Why Fixed-point
Floating point offers great range of values and more precision, but it also significantly more expensive when compared to integer or fixed point math.

With fixed point representation Embedded systems and Digital Signal processing designers can achieve greater speed and reduce the hardware design cost.

Note: Designer need to make decision on tolerance of quantization/truncation errors occur in fixed point implementation.