產品特色

• PGF77 native OpenMP and auto-parallel FORTRAN 77 compiler
• PGCC OpenMP and auto-parallel C++ and ANSI/K&R C compilers (C++ not currently available on Windows; ANSI C only)
• PGDBG OpenMP parallel graphical debugger (not currently available on Windows)
• PGPROF OpenMP parallel performance profiler (command-level only on Windows)
• Full 64-bit support on AMD Opteron, AMD Athlon64 and Intel Xeon-64 including full support for -mcmodel=medium and single data objects > 2GB (> 2GB objects currently not support in C++).
• AMD Opteron optimizations including SSE/SSE2, prefetching, use of extended register sets, and 64-bit addressing
• Pentium II/III/4/Xeon and AMD Athlon/AthlonXP optimizations including SSE/SSE2 and prefetching where supported in hardware
• Includes AMD Core Math Library (ACML) version 2.0
• Supported on 32-bit and 64-bit Linux and 32-bit Windows (NT/2000/XP)
• Parallel executables generated by PGI's Server-class compilers use up to 16 processors
• Tracking ANSI C++ Standard - EDG 3.30 C++ front-end
• ANSI and K&R C compilation modes
• Threads-based auto-parallelization of F77 DO loops
• Threads-based auto-parallelization of for loops
• Native parallelization using OpenMP directives in Fortran
• Native parallelization using OpenMP pragmas in C/C++
• Support for the SGI C$DOACROSS directive and "-mp" compile switch.
• Extensive vectorization / optimization directives support
• UNIX-compatible build/edit environment for Windows, including the BASH shell, vi and
• EMACS editors, make, tar, gzip, sed, grep, awk, and over 100 other shell commands!
• Large file (> 2GB) support in Fortran on 32-bit x86 systems
• -r8/-i8 compilation flags, 64-bit integers
• State-of-the-art dependence analysis and global optimization
• Vectorization / loop interchange / loop splitting
• Function inlining
• Loop unrolling
• Invariant conditional removal
• Memory hierarchy optimizations / cache tiling
• One pass interprocedural analysis (IPA)
• Byte-swapping I/O allows reads/writes of binary data files produced on Sun or SGI workstations
• Ability to link g77-compiled program units into PGF77 main programs
• Interoperable with Etnus TotalView®
• Integrated cpp pre-processing
• Compile-time optimization listings using the "-Minfo" switch
• Use of fast builtin transcendental functions in C/C++
• ANSI and K&R C compilation modes
• C++ Class member templates
• C++ partial specialization and ordering
• C++ explicit template qualification
• Support for creation of shared objects on Linux and DLL on Windows
• Cross-OS licensing at no extra charge -- use under any OS on multi-boot systems
• Unconditional 30 day money back guarantee