Skip to main content

Framework for functional hardware design approach

Project description

HW Design: A Functional Approach

PyGears is a free framework that lets you design hardware using high-level Python constructs and compile it to synthesizable SystemVerilog or Verilog code. There is a built-in simulator that lets you use arbitrary Python code with its vast set of libraries to verify your hardware modules. PyGears makes connecting modules easy, and has built-in synchronization mechanisms that help you build correct parallel systems.

@gear
def echo(samples: Fixp, *, feedback_gain, sample_rate, delay):

    sample_dly_len = round(sample_rate * delay)
    fifo_depth = ceil_pow2(sample_dly_len)
    feedback_gain_fixp = samples.dtype(feedback_gain)

    dout = Intf(samples.dtype)

    feedback = decouple(dout, depth=fifo_depth) \
        | prefill(dtype=samples.dtype, num=sample_dly_len)

    feedback_attenuated = (feedback * feedback_gain_fixp) \
        | samples.dtype

    dout |= (samples + feedback_attenuated) | samples.dtype

    return dout

Python functions model hardware modules, where function arguments represent module inputs and parameters. Example echo module has a single input port called samples where data of arbitrary signed fixed-point type Fixp can be received. Other three parameters feedback_gain, sample_rate and delay are compile time parameters.

@gear
def echo(samples: Fixp, *, feedback_gain, sample_rate, delay):
    ...

Arbitrary Python code can be used in modules at compile time, for an example to transform input parameters:

sample_dly_len = round(sample_rate * delay)
fifo_depth = ceil_pow2(sample_dly_len)
feedback_gain_fixp = samples.dtype(feedback_gain)

Rest of the echo function code describes the hardware module for applying echo audio effect to the input stream.

images/echo.png

Modules are instantiated using function calls: decouple(dout, depth=fifo_depth), which return module output interfaces that can in turn be passed as arguments to other module functions in order to make a connection between the modules. For conveniance the pipe "|" operator can be used to pass output of one function as argument to the next one. This was used to connect the output of decouple to prefill ("\" is used just to split the line visually):

feedback = decouple(dout, depth=fifo_depth) \
    | prefill(dtype=samples.dtype, num=sample_dly_len)

Again, the echo function returns its output interfaces which is then used to establish the connection with the next module that received the echo output stream:

@gear
def echo(...):
    ...
    return dout

Built-in simulator makes it easy to test and verify the modules while drawing power from the Python vast ecosystem of libraries. For an example, use Python built-in audioop library to read WAV files into an input samples stream for the echo module, and then visualise the input and output waveforms using matplotlib:

images/echo_plot.png

Speedup the simulation by configuring PyGears simulator to use open-source Verilator to simulate hardware modules, or some of the proprietary simulators like Questa, NCSim or Xsim. Implement any part of the system in a standard HDL and debug your design by inspecting the waveforms for an example in open-source wave viewer GTKWave

images/echo_vcd.png

Checkout Echo example description for more in depth information about the echo example.

Installation instructions

Install PyGears package with the Python package manager. On Linux distributions, depending on how your Python was installed you might get an error and need to prefix the command with sudo:

pip3 install pygears

For more detailed installation instructions (including how to install additional software) checkout Installation page.

Read the documentation

PyGears documentation

Checkout the examples

Library of standard modules

Echo: Hardware module that applies echo audio effect to a continuous audio stream.

RISC-V processor: PyGears implementation. Checkout also the RISC-V implementation blog series.

Tests: Contain many examples on how individual PyGears components operate

Contributions

Special thanks to the people that helped develop this framework:

  • Andrea Erdeljan

  • Damjan Rakanović

  • Nemanja Kajtez

  • Risto Pejašinović

  • Stefan Tambur

  • Vladimir Nikić

  • Vladimir Vrbaški

In order to contribute, pull your copy from github repository and create a pull request.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pygears-0.3.1.tar.gz (295.7 kB view hashes)

Uploaded Source

Built Distribution

pygears-0.3.1-py3-none-any.whl (449.3 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page