A gentle introduction to CocoTb
In this tutorial, we look at cocotb, an upcoming popular verification framework that let's you control signals directly from python!
Archive12 posts
Creating automated testbenches for your digital designs using python and iverilog
In this article, we look at a quick way to automate your testbenches and find bugs in your code.
Putting it all together: The CNN accelerator
In this article, we put all the modules that we've designed in this series together and build a single unit of a CNN architecture.
How to choose an FPGA dev board in 2020. A comprehensive guide
This comprehensive guide aims to guide you in your search for the perfect FPGA development board.
How to reset your FPGA design at start up without using an external pin or button
An article showing beginners how they can generate a global reset signal at the initialization of the design without having to sacrifice a GPIO pin or a user button for the same
Adding Fixed point arithmetic to your design
In this article we convert normal additions and multiplications in our design to fixed point representations. This will enable use to work with fractional numbers.
The Activation Function
In this article we implement some common activation functions used in machine learning algorithms. This is a part of the series where we design a Convolutional Neural Network accelerator targeted towards FPGAs
The Pooling Unit
In this article we design a module that can very efficiently perform Max Pooling and Average pooling of an input matrix in Verilog HDL. This will go into our Deep Learning accelerator as mentioned in the previous articles in the series.
FPGA based acceleration of machine learning algorithms involving convolutional neural networks
There has been a lot of interest in the tech community lately to accelerate data intensive machine learning inference operations. This series of articles goes into a great detail in implementing an FPGA based accelerator in verilog HDL.
Architecture Outline : Machine Learning Accelerator
There has been a lot of interest in the tech community lately to accelerate data intensive artificial intelligence inference operations. This series of articles goes into a great detail in implementing an FPGA based AI accelerator in Verilog HDL.
The convolution Engine
In this article, we design a fully parameterized 2D convolution engine in Verilog HDL and verify its functionality with a golden model written in python.
How to write a parameterized universal shift register in verilog
A Shift registers is a very common design element in any digital design. This will be the only shift register you will ever need to write. The parameters in this code can be modified to create a shift register of any size and bit-width.