#SIMPLY FORTRAN CUDA SUPPORT DRIVER#
#SIMPLY FORTRAN CUDA SUPPORT PLUS#

There are a few differences in how CUDA concepts are expressed usingįortran 90 constructs, but the programming model for both CUDA Fortran andĬUDA Fortran is essentially Fortran with a few extensions that allow one toĮxecute subroutines on the GPU by many threads in parallel. If you are familiar with CUDA C, then you are already well on your way to usingĬUDA Fortran is based on the CUDA C runtime API. Rendering, but are general enough to be useful in many cases involvingĭata-parallelism, linear algebra, and other common use cases in scientificĬUDA Fortran is the Fortran interface to the CUDA parallel computing platform. GPU designs are optimized for the kind of computations found in graphics Parallel computational units with very high memory bandwidth, and potential

#SIMPLY FORTRAN CUDA SUPPORT PLUS#

Plus 46 user-requested enhancements and fixes.Graphic processing units or GPUs have evolved into programmable, highly.

CUDA 9.2 and 10.0 are no longer included in the PGI packages.

Official support for the nonvvm sub-option -ta=tesla and -Mcuda dropped.

Dropped support for the macOS platform.

The PGI Debugger pgdbg has been discontinued.

#SIMPLY FORTRAN CUDA SUPPORT DRIVER#

The pgf77 driver is no longer provided.

Added support for recent Linux distributions:.

Upgraded the PGI C compiler pgcc significantly.

Improved implementation of the OpenMP API to follow the OpenMP specification.

Improved debugging metadata for select procedures.Upgraded all Linux compilers to use LLVM 9.0.Replaced the OpenMP runtime used by the default Linux LLVM compilers.Added support for the CUDA Toolkit version 10.2.Updated the profiling interface to conform to the OpenACC 3.0 specification.Dropped support for _CUDA_API_VERSION.Changed copy behavior for OpenACC reductions to adhere to the OpenACC.Implemented support for on_device in CUDA Fortran.Improved support for attributes(host,device) functions in CUDA Fortran.Added support for capture of the *this pointer by value to the OpenACC.Added support for calling additional built-in CUDA math functions.To the appropriate cuTENSOR functionality. New cutensorEx module maps select Fortran intrinsic functions.Added Fortran interfaces to the supplied NVIDIA cuTENSOR library.

See the PGI Release Notes for x86 and PGI Release Notes and Installation Guide for OpenPower for the complete list of new features.

Plus 29 user-requested enhancements and fixes.

Updated the implementation of acc_copyin and acc_create to adhere to the OpenACC 2.0 specification.

Default CUDA Toolkit version is now 10.2.

Added DWARF metadata for inlined function and subroutine calls.

Improvement performance in TRANSFER intrinsic for large contiguous data transfers.

C/C++ Compilers: Added DWARF metadata for inlined function calls.

Added support for interoperability with GNU-compiled OpenMP objects.

Bug fixes are listed on the PGI Release 2020 Technical Problem Reports summary page.Ĭomplete details on the new features listed here are included in the PGI Release Notes for x86-64 and PGI Installation Guide and Release Notes for OpenPOWER.

The new features added with each version are listed below. PGI releases version updates, as often as monthly, which may include both new features and bug fixes. A summary of new features in each version of the PGI 2020 release.

Simply fortran cuda support