Skip to content
Snippets Groups Projects
Commit 250b9c60 authored by jf3g19's avatar jf3g19
Browse files

Update README.md

parent 0fb2d703
Branches
No related tags found
No related merge requests found
**Contents**
Implementation of a Basic Distributed Arithmetic Architechture for MAC
**Operation**
- The Distributed Arithmetic Architechture is based on the assumption that for a given Sum Of Products (SoP) between two arrays `C` and `X` that you know the value of `C` at compile time.
- Based on this assumption we can generate a LUT which allows us to transform the SoP to a series of shifts and adds.
- The File generate_LUT.cpp can be used to generate the LUT values for a given set of input coefficients. If you compile the file it can be run in the command line in the following format:
`> generate_LUT [Multiplicand Bit-Width] [C[0]] [C[1]] [C[2]] [C[3]] .... [C[n]] `
i.e
`generate_LUT 16 -4 -3 1 9`
would generate the LUT for the coefficients c[0] = -4, c[1] = -3, c[2] = 1, c[3] = 9, and a 16-bit Bit-Width for your Multiplicand.
The generate multiplication coefficients can be stored as an unpacked array of packed values and should be connected directly to the `multiplication_coefficients` input port.
**Example**
An Example ModelSim output for the testbench provided, which displays the inner product of
`X = [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16]`
`C = [ [-79 -136 312 654 -1244 -2280 4501 14655 14655 4501 -2280 -1244 654 312 -136 -79]`
[output](https://imgur.com/WczvADF)
We can see that after input ready goes high, 17 clock cycles later, output ready goes high [with the correct value 278,511](https://imgur.com/ctePVPN)
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Please register or to comment