12.05.2018       Выпуск 229 (07.05.2018 - 13.05.2018)       Интересные проекты, инструменты, библиотеки

Chips 2.0 - переводим из C в Verilog с помощью Python Bindings


Экспериментальная функция:

Ниже вы видите текст статьи по ссылке. По нему можно быстро понять ссылка достойна прочтения или нет

Просим обратить внимание, что текст по ссылке и здесь может не совпадать.

https://travis-ci.org/dawsonjon/Chips-2.0.svg?branch=master https://readthedocs.org/projects/chips-20/badge/?version=latest

Chips - 2.0

Chips is a high level, FPGA design tool inspired by Python.

Try it out

Why not try the Chips web app.

Design components in C, design FPGAs in Python

In Chips, a design resembles a network of computers implemented in a single chip. A chip consists of many interconnected components operating in parallel. Each component acts like a computer running a C program.

Components communicate with each other sending messages across buses. The design of a chip - the components and the connections between them - is carried in Python.

Chips come in three parts:

  1. A Python library to build and simulate chips by connecting together digital components using high speed buses.
  2. A collection of ready made digital components.
  3. A C-to-hardware compiler to make new digital components in the C programming language.

A quick example

from chips.api.api import *

#create a new chip
chip = Chip("knight_rider")

#define a component in C
scanner = Component(C_file = """

    /* Knight Rider */
    int leds = output("leds");
    void main(){
        unsigned shifter = 1;
            while(shifter != 0x80){
                fputc(shifter, leds);
                shifter <<= 1;
            while(shifter != 0x01){
                fputc(shifter, leds);
                shifter >>= 1;

""", inline=True)

#capture simulation output in Python
scanner_output = Response(chip, "scanner", "int")

#add scanner to chip and connect
scanner(chip, inputs = {}, outputs = {"leds":scanner_output})

#generate synthesisable verilog code

#run simulation in Python
while len(scanner_output) < 16:

#check the results
print list(scanner_output)

Work at a higher level of abstraction

In Chips, the details of gates, clocks, resets, finite-state machines and flow-control are handled by the tool, this leaves the designer free to think about the architecture and the algorithms. This has some benefits:

  • Designs are simpler.
  • Simpler designs take much less time to get working.
  • Simpler designs are much less likely to have bugs.

With Chips the batteries are included

With traditional Hardware Description Languages, there are many restrictions on what can be translated into hardware and implemented in a chip.

With Chips almost all legal code can be translated into hardware. This includes division, single and double precision IEEE floating point, maths functions, trig-functions, timed waits, pseudo-random numbers and recursive function calls.

Python is a rich verification environment

Chips provides the ability to simulate designs natively in Python. Python is an excellent programming language with extensive libraries covering many application domains. This makes it the perfect environment to verify a chip.

NumPy , SciPy and MatPlotLib will be of interest to engineers, but that's just the start .

Under the hood

Behind the scenes, Chips uses some novel techniques to generate compact and efficient logic - a hybrid of software and hardware.

Not only does the compiler translate the C code into CPU instructions, it also generates a customised pipelined RISC CPU on the fly. The CPU provides the optimal instruction set for any particular C program.

By minimising the logic required to perform each concurrent task, designers can reduce power and area or cost. Performance gains can be achieved by increasing the number of concurrent tasks in a single device (tens in a small device to around a thousand or more large device).

While the code generated by chips is compact and efficient, die hard FPGA designers will be pleased to know that they can still hand craft performance critical data paths if they need to. There are even a few hand crafted components thrown in!

Install from github

$ git clone --recursive https://github.com/dawsonjon/Chips-2.0.git
$ cd Chips-2.0
$ sudo python setup install

Install from PyPi

$ pip-install chips

Лучшая Python рассылка

Разместим вашу рекламу

Пиши: mail@pythondigest.ru

Нашли опечатку?

Выделите фрагмент и отправьте нажатием Ctrl+Enter.

Система Orphus