Python 3.7.4 : About with the PyOpenCL python module.

PyOpenCL lets you access GPUs and other massively parallel compute devices from Python.
It is important to note that OpenCL is not restricted to GPUs.
In fact, no special hardware is required to use OpenCL for computation–your existing CPU is enough.
The documentation of this project can be found at this website.
Let's install the python module for python 3 version:

[mythcat@desk ~]$ pip3 install pyopencl --user
Collecting pyopencl
...
Successfully built pytools
Installing collected packages: pytools, pyopencl
Successfully installed pyopencl-2019.1.1 pytools-2019.1.1

The install of OpenCL driver can be done with these commands:

# get OpenCL driver automated installer (installs kernel 4.7)
curl https://software.intel.com/sites/default/files/managed/f6/77/install_OCL_driver.sh_.txt > install_OCL\
_driver.sh
chmod +x install_OCL_driver.sh
# install OpenCL driver
sudo ./install_OCL_driver.sh install
# check
ls /boot/vmlinuz-*intel*

This is a simple python script to test the opencl context:

import pyopencl as cl
import numpy as np
ctx = cl.create_some_context()
# cet platforms, both CPU and GPU
my_plat= cl.get_platforms()
CPU = my_plat[0].get_devices()
try:
    GPU = my_plat[1].get_devices()
except IndexError:
    GPU = "none"
# create context for GPU/CPU
if GPU != "none":
    ctx = cl.Context(GPU)
else:
    ctx = cl.Context(CPU)
# create queue for each kernel execution
queue = cl.CommandQueue(ctx)
mf = cl.mem_flags

This is another simple python script:

# -*- coding: utf-8 -*-
import pyopencl as cl 
import numpy
a = numpy.random.rand(50000).astype(numpy.float32)
ctx = cl.create_some_context()
queue = cl.CommandQueue(ctx)
a_buf = cl.Buffer(ctx ,cl.mem_flags.READ_WRITE,size=a.nbytes)
cl.enqueue_write_buffer(queue, a_buf , a)

prg= cl.Program(ctx,
"""
__kernel void twice(__global float ∗a)
{
 int gid=get_global_id(0);
 a[gid] ∗= 2;
}"""
). build()

prg.twice(queue, a.shape, None,a_buf ).wait()