Installation¶
NumbaPro is part of the Anaconda Accelerate product. Please refer to the Anaconda Accelerate Documentation for instructions on downloading and installation.
Purchasing a License¶
Accelerate is included with Anaconda Workgroup and Anaconda Enterprise subscriptions.
To start a 30-day free trial just download and install the Anaconda Accelerate package.
If you already have Anaconda (free Python distribution) installed:
conda update conda
conda install accelerate
If you do not have Anaconda installed, you can download it here.
Anaconda Accelerate can also be installed into your own (non-Anaconda) Python environment. For more information about Accelerate please contact sales@continuum.io.
Update Instructions¶
With Anaconda Accelerate already installed, first update the conda package management tool to the latest version, then use conda to update the NumbaPro module. In the terminal:
$ conda update conda
$ conda update numbapro
CUDA GPUs Setup¶
CUDA Driver¶
NumbaPro does not ship the CUDA driver. It is users responsibility to ensure their systems are using the latest driver. Currently, users should use the driver shipped with CUDA 5.5 SDK.
CUDA Support & Detection¶
NumbaPro GPU support currently requires NVIDIA CUDA GPUs with compute-capability 2.0 or above. Users should check their hardware with the following:
import numbapro
numbapro.check_cuda()
A sample output looks like:
------------------------------libraries detection-------------------------------
Finding cublas
located at /Users/.../lib/libcublas.dylib
Finding cusparse
located at /Users/.../lib/libcusparse.dylib
Finding cufft
located at /Users/.../lib/libcufft.dylib
Finding curand
located at /Users/.../lib/libcurand.dylib
Finding nvvm
located at /Users/.../lib/libnvvm.dylib
In /Users/.../lib
finding libdevice.compute_20.bc ok
finding libdevice.compute_30.bc ok
finding libdevice.compute_35.bc ok
-------------------------------hardware detection-------------------------------
Found 1 CUDA devices
id 0 GeForce GT 650M [SUPPORTED]
compute capability: 3.0
pci device id: 0
pci bus id: 1
Summary:
1/1 devices are supported
This performs CUDA library and GPU detection. Discovered GPUs are listed with information for compute capability and whether it is supported by NumbaPro.
The message “cuda disabled by user” means that either the environment variable NUMBA_DISABLE_CUDA is set to 1 and must be set to 0, or the system is 32-bit. CUDA does not support 32-bit systems.