created: 05/27/2002; revisions:11/06/02, 11/26/03; 05/24/2015

Up until now all of the arithmetic we have done on the MIPS has been with integers, both signed (two's complement) and unsigned. This chapter starts the discussion of how floating point arithmetic is done on the MIPS.

- Positional notation with fractions
- Converting fractions between base two and base ten
- Fixed point representation
- Limited precision of binary fractions (fixed point and floating point)
- How "one tenth" cannot be expressed in binary
- Dangerous program loops

Is is possible to represent a
**negative** integer with a **positive**
number of bits?