improved installation instructions

README.rst

@@ -63,7 +63,7 @@ Features
     $ mpiexec -n [nodes] python <your_ptypy_script>.py
 
 * **GPU acceleration** based on custom kernels, CuPy or PyCUDA/reikna.
-  See examples in ``templates/accelerate``. 
+  See examples in ``templates/accelerate``, ``templates/engines/pycuda`` and ``templates/engines/cupy``.
 
 * A **client-server** approach for visualization and control based on 
   `ZeroMQ <http://www.zeromq.org>`_ .
@@ -104,7 +104,7 @@ Ptypy depends on standard python packages:
 GPU support
 -----------
 
-We support an accelerated version of |ptypy|_ for CUDA-capable GPUs based on our own kernels and the
+We support an accelerated version of |ptypy| for CUDA-capable GPUs based on our own kernels and the
 `CuPy <https://cupy.dev/>`_ package. We recommend to install the dependencies for this version like so.
 ::
 

doc/index.rst

@@ -36,7 +36,7 @@ Highlights
     $ mpiexec/mpirun -n [nodes] python <your_ptypy_script>.py
 
 * **GPU acceleration** based on custom kernels, CuPy or PyCUDA/reikna.
-  See examples in ``templates/accelerate``. 
+  See examples in ``templates/accelerate``, ``templates/engines/pycuda`` and ``templates/engines/cupy``.
 
 * A **client-server** approach for visualization and control based on 
   `ZeroMQ <http://www.zeromq.org>`_ .

doc/rst_templates/getting_started.tmp

@@ -101,20 +101,6 @@ Install the dependencies for this version like so.
     (ptypy_cupy)$ pip install .
 
 
-While we use the `CuPy FFT <https://docs.cupy.dev/en/stable/reference/fft.html>`_ to
-provide a filtered FFT, i.e. a FFT that is fused with a pointwise
-matrix multiplication, you can optionally also install a version
-based on cufft and callbacks. Due to the nature of this extension it
-needs to be built for fixed array sizes externally. The currently supported 
-array sizes are 16, 32, 64, 128, 256, 512, 1024 and 2048.
-::
-
-    $ conda activate ptypy_cupy
-    (ptypy_cupy)$ cd cufft
-    (ptypy_cupy)$ conda env update --file dependencies.yml --name ptypy_cupy
-    (ptypy_cupy)$ pip install .
-
-
 Install for GPU support with PyCUDA
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
@@ -130,12 +116,16 @@ Install the dependencies for this version like so.
     (ptypy_pycuda)$ pip install .
 
 
-While we use `Reikna <https://pypi.org/project/reikna/>`_ to
-provide a filtered FFT, i.e. a FFT that is fused with a pointwise
-matrix multiplication, you can optionally also install a version
-based on cufft and callbacks. Due to the nature of this extension it
-needs to be built for fixed array sizes externally. The currently supported 
-array sizes are 16, 32, 64, 128, 256, 512, 1024 and 2048.
+We use `Reikna <https://pypi.org/project/reikna/>`_ to
+provide a filtered FFT, i.e. a FFT that is fused with a pointwise matrix multiplication. 
+
+Optional installation of filtered cufft
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+For optimal performance with both PyCUDA and CuPY engines, you can optionally install a version 
+of the filtered FFT based on cufft and callbacks. Due to the nature of this extension it
+needs to be built for fixed array sizes externally and currently supports array 
+sizes of 16, 32, 64, 128, 256, 512, 1024 and 2048.
 ::
 
     $ conda activate ptypy_pycuda