CODEWITHSUNDEEP
vhdldivisionfpga
xx77aBs
xx77aBs

Reputation: 4768

Easiest way to do VHDL floating point division?

I'm looking for easiest way to divide two floating point numbers using VHDL. I need the code to be synthesizable (I'll be implementing it on Spartan 3 FPGA).

First operand will always be a fixed number (e.g. 600), and second one will be integer, let's say between 0 and 99999. Fixed number is dividend, and the integer one is divisor. So I'll have to calculate something like this: 600/124. Or any other number instead of 124, of course that is in range between 0 and 99999. Second number (the one that is changing) will always be integer !! (there won't be something like 123.45).

After division, I need to convert the result into integer (round it up or just ignore numbers after decimal point, which ever is faster).

Any ideas ? Thanks !

Upvotes: 1

Views: 5197

Answers (2)

user597225
user597225

Reputation:

There are many ways to do this, with the easiest being a ROM. You don't need floating point anywhere since doing an integer divide and compensating for a non-zero remainder can give you the same results. I'd suggest calculating the first 600 results in MATLAB or a spreadsheet so you can see that handling values up to 99999 isn't necessary.

Also, some common nomenclature for range and precision is QI.F where I is the number of integer bits and F is the number of fractional bits. Thus 0..99999 would be Q17.0 and your output would be Q10.0.

Upvotes: 3

David Pointer
David Pointer

Reputation: 935

There's an FP divide function in this VHDL file from this site.

Upvotes: 3

Related Questions