ArrayFire v3.4.2 Release Notes
Release Date: 2016-12-21 // over 7 years ago-
v3.4.2
The source code with submodules can be downloaded directly from the following link:
http://arrayfire.com/arrayfire_source/arrayfire-full-3.4.2.tar.bz2Installer CUDA Version: 8.0 (Required)
Installer OpenCL Version: 1.2 (Minimum)๐ Deprecation Announcement
๐ This release supports CUDA 6.5 and higher. The next ArrayFire release will
๐ support CUDA 7.0 and higher, dropping support for CUDA 6.5. Reasons for no
๐ longer supporting CUDA 6.5 include:- ๐ CUDA 7.0 NVCC supports the C++11 standard (whereas CUDA 6.5 does not), which
is used by ArrayFire's CPU and OpenCL backends. - Very few ArrayFire users still use CUDA 6.5.
๐ As a result, the older Jetson TK1 / Tegra K1 will no longer be supported in
๐ the next ArrayFire release. The newer Jetson TX1 / Tegra X1 will continue to
have full capability with ArrayFire.๐ณ Docker
๐ Improvements
- Implemented sparse storage format conversions between AF_STORAGE_CSR
and AF_STORAGE_COO.
1- Directly convert between AF_STORAGE_COO <--> AF_STORAGE_CSR
๐ using the af::sparseConvertTo() function. - af::sparseConvertTo() now also supports converting to dense.
- Directly convert between AF_STORAGE_COO <--> AF_STORAGE_CSR
- ๐ Added cast support for sparse arrays.
1- Casting only changes the values array and the type. The row and column
index arrays are not changed.
- Casting only changes the values array and the type. The row and column
- Reintroduced automated computation of chart axes limits for graphics functions.
1- The axes limits will always be the minimum/maximum of the current and new
limit. - The user can still set limits from API calls. If the user sets a limit
from the API call, then the automatic limit setting will be disabled.
- The axes limits will always be the minimum/maximum of the current and new
- Using
boost::scoped_array
instead ofboost::scoped_ptr
when managing
array resources.
1 - ๐ Internal performance improvements to getInfo() by using
const
references
to avoid unnecessary copying ofArrayInfo
objects.
1 - โ Added support for scalar af::array inputs for af::convolve() and
set functions.
1
2
3 - ๐ Performance fixes in af::fftConvolve() kernels.
1
2
๐ Build
- ๐ Support for Visual Studio 2015 compilation.
1
2 - ๐ Fixed
FindCBLAS.cmake
when PkgConfig is used.
1
๐ Bug fixes
- ๐ Fixes to JIT when tree is large.
1
2 - ๐ Fixed indexing bug when converting dense to sparse af::array as
AF_STORAGE_COO.
1 - ๐ Fixed af::bilateral() OpenCL kernel compilation on OS X.
1 - ๐ Fixed memory leak in af::regions() (CPU) and af::rgb2ycbcr().
1
2
3
Installers
- ๐ Major OS X installer fixes.
1- Fixed installation scripts.
- Fixed installation symlinks for libraries.
- ๐ Windows installer now ships with more pre-built examples.
Examples
- โ Added af::choleskyInPlace() calls to
cholesky.cpp
example.
1
๐ Documentation
- โ Added
u8
as supported data type ingetting_started.md
.
1 - ๐ Fixed typos.
1
CUDA 8 on OSX
- ๐ CUDA 8.0.55 supports Xcode 8.
1
Known Issues
- Known failures with CUDA 6.5. These include all functions that use
๐ sorting. As a result, sparse storage format conversion between
AF_STORAGE_COO and AF_STORAGE_CSR has been disabled for CUDA 6.5.
- ๐ CUDA 7.0 NVCC supports the C++11 standard (whereas CUDA 6.5 does not), which