Hi all,
Apologies for posting across lists; I thought that this might be of
interest to both groups.
I have just released PyViennaCL 1.0.0, which is a set of largely
NumPy-compatible Python bindings to the ViennaCL linear algebra and
numerical computation library for GPGPU and heterogeneous systems.
PyViennaCL aims to make powerful GPGPU computing really transparently
easy, especially for users already using NumPy for representing
matrices.
Please see my announcement below for links to source and packages and
documentation, a list of features, and a list of missing pieces. I hope
to iron out all those missing bits over the coming months, and work on
closer integration, especially with PyOpenCL / PyCUDA, over the summer.
Best wishes,
Toby St Clere Smithe
--- Begin Message ---
Dear ViennaCL users,
If you've ever used Python for your numerical applications, you know
what joy it can be. Now, the easy power of ViennaCL 1.5.1 is at last
married to that experience. I am pleased to announce the first release
of PyViennaCL!
Download links for source and Ubuntu binaries are found at the usual
place: http://viennacl.sourceforge.net/viennacl-download.html
* If you are or know anyone who could help with building PyViennaCL for
other systems (Windows, Mac OS X, CentOS / RHEL, Fedora, SuSE, ...),
please get in touch!
See the following link for documentation and example code:
http://viennacl.sourceforge.net/pyviennacl/doc/
PyViennaCL 1.0.0 exposes most of the functionality of ViennaCL:
+ sparse (compressed, co-ordinate, ELL, and hybrid) and dense
(row-major and column-major) matrices, vectors and scalars on your
compute device using OpenCL;
+ standard arithmetic operations and mathematical functions;
+ fast matrix products for sparse and dense matrices, and inner and
outer products for vectors;
+ direct solvers for dense triangular systems;
+ iterative solvers for sparse and dense systems, using the BiCGStab,
CG, and GMRES algorithms;
+ iterative algorithms for eigenvalue estimation problems.
PyViennaCL has also been designed for straightforward use in the context
of NumPy and SciPy: PyViennaCL objects can be constructed using NumPy
arrays, and arithmetic operations and comparisons in PyViennaCL are
type-agnostic.
Some ViennaCL functionality is not yet available, and these features are
planned for a release in the coming months:
+ preconditioners and QR factorization;
+ additional solvers and other algorithms, such as FFT computation;
+ structured matrices;
+ CUDA support (use OpenCL for now!);
+ advanced OpenCL integration.
Spread the word!
Toby St Clere Smithe
--- End Message ---
_______________________________________________
NumPy-Discussion mailing list
NumPy-Discussion@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion
