Jacket 1.2.1 r3139: Development Release

[Old posts from the commercial version of ArrayFire] Discussion of ArrayFire using CUDA or OpenCL.

Moderator: pavanky

Jacket 1.2.1 r3139: Development Release

Postby vishy » Mon Dec 28, 2009 4:02 pm

We have released an in-development version of Jacket 1.2.1 in order to let you sample the fixes and new features we have incorporated.
Some features are:
- Introduction of a newer version of the LAPACK Library, CULA 1.0 Basic
- All-new features added to the Graphics Toolbox
- Support for functions such as besselj, besselh and more

You are free to try out this development version, keeping in mind that it contains some issues that are yet to be resolved. Some of these issues are outlined below:
- This development version requires Linux users to set LD_LIBRARY_PATH in the following order: <PATH TO JACKET>/cula/lib:<PATH TO CUDA>/lib
(Replace lib by lib64 if using 64-bit Linux)
- This version of Jacket has an issue with the memory subsystem that will be fixed in the next version. This may manifest itself in the GPU issuing out of memory errors on Windows 7.

As always, we welcome all feedback and suggestions for improvement of Jacket. Please use the forums or email us at support@accelereyes.com if you have problems installing or using this development version.

This version of Jacket can be downloaded for Windows (32- and 64-bit) and Linux (32- and 64-bit) from the following locations:
Win32 and Win64: http://www.accelereyes.com/special/Jack ... 091224.exe
Linux32 and Linux64: http://www.accelereyes.com/special/Jack ... 091224.run


The following are the Release Notes for this dev version:

* Requires CUDA 2.3 Drivers and Toolkit

Additions:
+ reallog, realsqrt now fully supported
+ besselj, besselh supported for gsingles
+ Arithmetic functions: factorial, fix, kron, cross
+ Signal Processing: filtfilt

Changes:
+ gones, gzeros, geye can now return double-precision by default on double-precision
cards. This is possible by setting a flag as follows: gpu_entry(10,true) to
enable default-gdouble, and gpu_entry(10,false) to disable it.
+ gones,gzeros,geye,grand and grandn (without arguments) now return a 1x1 number.
+ sqrt and power now supported for complex inputs
+ Optimizations made to subscripted referencing and assignments
+ The divide operation now preserves class: if A is a gdouble, 1/A has the
same class
+ Fixed issue with incorrect cached kernel being used in multiplication
+ Fixed issue where mldivide failed with CPU arguments
+ Fixed failures with complex gdouble to complex gsingle conversion
+ Fixed failures with mod, blkdiag for some inputs
+ Fixed a failure for cat(1,A,B) for 2D inputs (order of 1000x500).

Known issues:
+ If you are running Jacket in Linux on 'screen', note that problems may occur
with commands such as 'ginfo' and 'gactivate'. This is because Jacket sets
the environment variable KMP_DUPLICATE_LIB_OK to TRUE to enable the SVD
algorithm to run.
+ bsxfun, cov, std, var, corrcoef currently give incorrect results for some
data (Please see release notes for 1.2). This is slated to be fixed soon.
--------------------------------------------------
Vish Venugopalakrishnan
Software Engineer (Q/A)
AccelerEyes LLC
vishy.v@accelereyes.com

--------------------------------------------------
Resources:
Getting Started, FAQ, Tips, Syntax
User avatar
vishy
 
Posts: 411
Joined: Thu Apr 16, 2009 11:46 am

Return to [archive-commercial] Programming & Development with ArrayFire

cron