Skip to main content
Warning: You are using the test version of PyPI. This is a pre-production deployment of Warehouse. Changes made here affect the production instance of TestPyPI (
Help us improve Python packaging - Donate today!

OpenCL-backed neural simulations using the methods of the Neural Engineering Framework.

Project Description

This project is an OpenCL-based simulator for brain models built using Nengo. It can be orders of magnitude faster than the reference simulator in nengo for large models.


To use the nengo_ocl project’s OpenCL simulator, build a nengo model as usual, but pass sim_ocl.Simulator when creating a simulator for your model:

import numpy as np
import nengo

# define the model
model = nengo.Network()
with model:
    stim = nengo.Node(np.sin)
    a = nengo.Ensemble(n_neurons=100, dimensions=1)
    b = nengo.Ensemble(n_neurons=100, dimensions=1)
    nengo.Connection(stim, a)
    nengo.Connection(a, b, function=lambda x: x**2, synapse=0.01)

    probe_a = nengo.Probe(a, synapse=0.01)
    probe_b = nengo.Probe(b, synapse=0.01)

from nengo_ocl.sim_ocl
import pyopencl as cl
ctx = cl.create_some_context()

# build the model
sim = nengo_ocl.sim_ocl..Simulator(model, context=ctx)
# run the model

# plot the results
import matplotlib.pyplot as plt

Dependencies and Installation

General: * Python 2.6 or better * One or more OpenCL implementations (test with e.g. PyOpenCl)

A working installation of OpenCL is the most difficult part of installing Nengo OCL. See below for more details on how to install OpenCL.

Python packages: * NumPy * nengo * networkx * mako * PyOpenCL

In the ideal case, all of the Python dependencies will be automatically installed when installing nengo_ocl with

pip install nengo_ocl

If that doesn’t work, then do a developer install to figure out what’s going wrong.

Developer Installation

First, pip install nengo. For best performance, make sure a fast version of Numpy is installed by following the instructions in the Nengo README. Currently, nengo_ocl is compatible with Nengo 2.0.x, supporting most features.

Once Nengo is installed, install the remaining dependencies:

pip install networkx mako pyopencl

This repository can then be installed with:

git clone
cd nengo_ocl
python develop --user

If you’re using a virtualenv (recommended!) then you can omit the --user flag.

Installing OpenCL

How you install OpenCL is dependent on your hardware and operating system. A good resource for various cases is found in the PyOpenCL documentation:

Below are instructions that have worked for the Nengo OCL developers at one point in time.

Intel OCL on Debian/Ubuntu Linux

Intel provides an OpenCL driver for at least some of their multicore processors. Core-i7 and Xeon chips can be quite good for running Nengo simulations.


  1. Download Intel SDK for OpenCL for applications from Intel’s OpenCL website

  2. Extract

    tar zxvf intel_sdk_for_ocl_applications_2012_x64.tgz
  3. Convert RPM files to .deb

    sudo apt-get install -y rpm alien libnuma1  # Get conversion packages
    fakeroot alien --to-deb opencl-1.2-*.rpm  # Convert all RPMs
  4. Install .deb packages. They will be put in /opt/intel

    sudo dpkg -i opencl-1.2-*.deb # Install all .debs
  5. Add library to search path

      sudo touch /etc/
    Put in the line: ``/opt/intel/opencl-1.2-3.0.67279/lib64``
  6. Link the Intel ICD file

    sudo ln /opt/intel/opencl-1.2-3.0.67279/etc/intel64.icd /etc/OpenCL/vendors/intel64.icd
  7. Run ldconfig

    sudo ldconfig

AMD OCL on Debian Unstable

On Debian unstable (sid) there are packages in non-free and contrib to install AMD’s OCL implementation easily. Actually, the easiest thing would be to apt-get install python-pyopencl. But if you’re using a virtual environment, you can sudo apt-get install opencl-headers libboost-python-dev amd-opencl-icd amd-libopencl1 and then pip install pyopencl.

Nvidia OCL on Debian/Ubuntu Linux

On Debian unstable (sid) there are packages for installing the Nvidia OpenCL implementation as well.

sudo apt-get install nvidia-opencl-common nvidia-libopencl1

Ensure that the Nvidia driver version matches the OpenCL library version. You can check the Nvidia driver version by running nvidia-smi in the command line. You can find the OpenCL library version by looking at the file in the /usr/lib/x86_64-linux-gnu/ folder.

Note! At the time of writing (Sept 2013) these drivers provide only OpenCL-1.1 rather than the more current OpenCL-1.2. Consequently, you may find that pyopencl’s default build creates a binary Python module ( that cannot be loaded (i.e. import pyopencl fails in the Python interpreter). You can fix this one of two ways:

  1. Use the generic driver-loading library from another provider (by e.g. following the Intel instructions above), and simply don’t try to use new 1.2 features on NVidia devices.
  2. Follow PyOpenCL’s build instructions to compile an OpenCL-1.1 version of PyOpenCL.

It’s nice to have a CPU OpenCL driver, so we recommend option (1).

Release History

0.1.0 (unreleased)

Initial release of Nengo OpenCL! Supports Nengo 2.0.x on Python 2.6+ and 3.3+.

Release History

Release History

This version
History Node


Supported By

WebFaction WebFaction Technical Writing Elastic Elastic Search Pingdom Pingdom Monitoring Dyn Dyn DNS Sentry Sentry Error Logging CloudAMQP CloudAMQP RabbitMQ Heroku Heroku PaaS Kabu Creative Kabu Creative UX & Design Fastly Fastly CDN DigiCert DigiCert EV Certificate Rackspace Rackspace Cloud Servers DreamHost DreamHost Log Hosting