@@ -3,17 +3,26 @@ Pre-built binaries are available on our [releases page](https://github.com/clMat
33
44| Build branch | master | develop |
55| -----| -----| -----|
6- | Linux/OSX x64 | [ ![ Build Status] ( https://travis-ci.org/clMathLibraries/clSPARSE.svg?branch=master )] ( https://travis-ci.org/clMathLibraries/clSPARSE ) | [ ![ Build Status] ( https://travis-ci.org/clMathLibraries/clSPARSE.svg?branch=develop )] ( https://travis-ci.org/clMathLibraries/clSPARSE ) |
6+ | GCC/Clang x64 | [ ![ Build Status] ( https://travis-ci.org/clMathLibraries/clSPARSE.svg?branch=master )] ( https://travis-ci.org/clMathLibraries/clSPARSE/branches ) | [ ![ Build Status] ( https://travis-ci.org/clMathLibraries/clSPARSE.svg?branch=develop )] ( https://travis-ci.org/clMathLibraries/clSPARSE/branches ) |
7+ | Visual Studio x64 | [ ![ Build status] ( https://ci.appveyor.com/api/projects/status/93518qe0efy6n7fy/branch/master?svg=true )] ( https://ci.appveyor.com/project/kknox/clsparse-otonj/branch/master ) | [ ![ Build status] ( https://ci.appveyor.com/api/projects/status/93518qe0efy6n7fy/branch/develop?svg=true )] ( https://ci.appveyor.com/project/kknox/clsparse-otonj/branch/develop ) |
78
89# clSPARSE
9- an OpenCL© ; library implementing Sparse linear algebra. This project is a result of
10+ an OpenCL&trade ; library implementing Sparse linear algebra routines . This project is a result of
1011a collaboration between [ AMD Inc.] ( http://www.amd.com/ ) and
1112[ Vratis Ltd.] ( http://www.vratis.com/ ) .
1213
13- ## Introduction to clSPARSE
14- At this time, clSPARSE provides these fundamental sparse operations for OpenCL:
14+ ### What's new in clSPARSE ** v0.8**
15+ - New single precision SpM-SpM (SpGEMM) function
16+ - Optimizations to the sparse matrix conversion routines
17+ - [ API documentation] ( http://clmathlibraries.github.io/clSPARSE/ ) available
18+ - SpM-dV routines now provide [ higher precision accuracy] (https://github.com/clMathLibraries/clSPARSE/wiki/Precision )
19+ - Various bug fixes integrated
20+
21+
22+ ## clSPARSE features
1523- Sparse Matrix - dense Vector multiply (SpM-dV)
1624- Sparse Matrix - dense Matrix multiply (SpM-dM)
25+ - Sparse Matrix - Sparse Matrix multiply Sparse Matrix Multiply(SpGEMM) - Single Precision
1726- Iterative conjugate gradient solver (CG)
1827- Iterative biconjugate gradient stabilized solver (BiCGStab)
1928- Dense to CSR conversions (& converse)
@@ -25,11 +34,7 @@ projects to build wrappers around clSPARSE in any language they need. A great d
2534of thought and effort went into designing the API’s to make them less ‘cluttered’
2635compared to the older clMath libraries. OpenCL state is not explicitly passed
2736through the API, which enables the library to be forward compatible when users are
28- ready to switch from OpenCL 1.2 to OpenCL 2.0 _ (OpenCL 2.0 support not completely finished)_
29-
30- The API’s are designed such that users are in control of where input and output
31- buffers live, and they maintain control of when data transfers to/from device
32- memory happen, so that there are no performance surprises.
37+ ready to switch from OpenCL 1.2 to OpenCL 2.0 <sup >[ 1] ( #opencl-2 ) </sup >
3338
3439### Google Groups
3540Two mailing lists have been created for the clMath projects:
@@ -57,15 +62,15 @@ script for clSPARSE also builds the samples as an external project, to demonstra
5762how an application would find and link to clSPARSE with cmake.
5863
5964### clSPARSE library documentation
60- ** API documentation** is not yet available, but the samples above give an excellent
65+ ** API documentation** is now available http://clmathlibraries.github.io/clSPARSE/ . The included samples will give an excellent
6166starting point to basic library operations.
6267
6368### Contributing code
6469Please refer to and read the [ Contributing] ( CONTRIBUTING.md ) document for guidelines on
6570how to contribute code to this open source project. Code in the
6671/master branch is considered to be stable and new library releases are made
6772when commits are merged into /master. Active development and pull-requests should
68- be made to the ** / develop** branch.
73+ be made to the ** develop** branch.
6974
7075## Build
7176clSPARSE is primarily written with C++ using C++11 core features. It does export
@@ -101,3 +106,6 @@ files
101106### Bench & Test infrastructure dependencies
102107- Googletest v1.7
103108- Boost v1.58
109+
110+ ## Clarifications
111+ <a name =" opencl-2 " >[ 1] </a >: OpenCL 2.0 support is not yet fully implemented; only the interfaces have been designed
0 commit comments