This package aims to provide GPU accelerated implementations of common volume processing algorithms to the python ecosystem, such as

• convolutions
• denoising
• synthetic noise
• ffts (simple wrapper around reikna)
• affine transforms

via OpenCL and the excellent pyopencl bindings.

Some examples of processing tasks and their respective runtime (tests/benchmark/benchmark.py):

[1] Xeon(R) CPU E5-2630 v4 using numpy/scipy functions
[2] NVidia Titan X using gputools
[3] as [2] but without CPU->GPU->CPU transfer

• python 2.7 / 3.5+
• a working OpenCL environment (check with clinfo).

pip install gputools

Or the developmental version:

pip install git+https://github.com/maweigert/gputools@develop

Check if basic stuff is working:

python -m gputools

Docs are still to be done ;)

Most of the methods work on both numpy arrays or GPU memory objects (gputools.OCLArrays/OCLImage). The latter saving the memory transfer (which e.g. for simple convolutions accounts for the main run time)

• 2D-3D convolutions
• separable convolutions
• fft based convolution
• spatially varying convolutions

import gputools

d = np.zeros((128,128), np.float32)
d[64,64] = 1.
h = np.ones((17,17))
res = gputools.convolve(d,h)

d = np.zeros((128,128,128), np.float32)
d[64,64,64] = 1.
hx,hy,hz = np.ones(7),np.ones(9),np.ones(11)
res = gputools.convolve_sep3(d,hx,hy,hz)

bilateral filter, non local means

...
d = np.zeros((128,128,128, np.float32))
d[50:78,50:78,50:78:2] = 4.
d = d+np.random.normal(0,1,d.shape)
res_nlm = gputools.denoise.nlm3(d,2.,2,3)
res_bilat = gputools.denoise.bilateral3(d,3,4.)

fast 2d and 3d perlin noise calculations

gputools.perlin3(size = (256,256,256), scale = (10.,10.,10.))

scaling, translate, rotate, affine...

gputools.transforms.scale(d,.2)
gputools.transforms.rotate(d,axis = (64,64,64),angle = .3)
gputools.transforms.shift(d,(10,20,30))
...

wraps around reikna

gputools.fft(d)
gputools.fft(d, inverse = True)

Some configuration data (e.g. the default OpenCL platform and devic) can be changed in the config file "~/.gputools" (create it if necessary)

#~/.gputools

id_platform = 0
id_device = 1

See

gputools.config.defaults

for available keys and their defaults.

Alternatively, the used OpenCL Device can be set via the environment variables gputools_id_device, gputools_id_platform, and gputools_use_gpu (variables present in the config file will take precendence, however).

If you see a

ImportError: _cffi.so: undefined symbol: clSVMFree

after importing gputools, this is most likely a problem of pyopencl being installed with an incorrent OpenCL version. Check the OpenCL version for your GPU with clinfo (e.g. 1.2):

clinfo | grep Version

and install pyopencl manually while enforcing your opencl version:

# uninstall pyopencl
pip uninstall pyopencl cffi
  
# get pyopencl source
git clone https://github.com/pyopencl/pyopencl.git
cd pyopencl
python configure.py
	
# add in siteconf.py the line
# CL_PRETEND_VERSION = "1.2"
echo 'CL_PRETEND_VERSION = "1.2"' >> siteconf.py

pip install .

where e.g. "1.2" is your version of OpenCL.