A simple floating point conversion tool that allows for an arbitrary size float. It also gives useful information about min/max values.
I have not done too much testing so I would still double check the output, specifically the decimal representations.
Instructions
This is IEEE 754 compatible as far as I know. (32 bit floating point value with 1 sign bit, 8 exponent bits, and 23 mantissa bits)
The binary values are specified as (1 bit sign) (x bits exponent) (y bits mantissa)
Everything is given as a magnitude, so simply negate a value if you need, say, the smallest possible normalized value.
Notes
The 1 sign bit is assumed.
Please note that the maximum supported exponent bit width is about 53. This shouldn't be an issue as even the octuple precision floating-point format only has 19 exponent bits.
- Type - | - Scientific notation (Binary) - | - Binary - | - Scientific notation (Decimal) - |
---|---|---|---|
Smallest normalized value | |||
Largest normalized value | |||
Smallest denormalized value | |||
Largest denormalized value | |||
Input - binary to decimal | not implemented | obvious | |
Input - decimal to binary | not implemented | not implemented | not implemented |
Contact me at [my first name]h@utexas.edu