csnotes/337/lec/lec17.md
2019-09-24 11:34:35 -07:00

30 lines
411 B
Markdown

# floating point
mips uses ieee754
## parts of floats
1. Sign bit
2. Exponent
3. Mantissa
_32-bit floats btw_
## Sign
bruh
## exponent
This is coming from scientific notation so we have `2.0E3`
This means our 8 bit exponent is `0000 0011` + `0111 1111`
## Mantissa
let's roll with `2.0E3`
Mantiss of 2.0 = 0 <- 23 times
We have 23 `0`'s because the decimal portion of our number `2.0E3` is 0.