Compilers and Languages for HPC

Using compilers on compute nodes

Compute nodes are diskless. Therefore the OS image is kept small, and much of the programming environment (like devel packages and headers) is missing there. Usually, build your software on login nodes!
If you still wish to compile/build software on compute nodes, you can try to load modules to setup your build environment. Still, we cannot warrant this to work under all circumstances!

Programming Languages
Language	Short description
C, C++	C is a versatile language targetted at systems programming, but also used for scientific data processing. C++ defines an extension of C that is more application-centric due to availability of powerful object-oriented, generic and other programming paradigms.
C, C++	Intel Compiler, GNU GCC, Portland Group Compilers, LLVM
Fortran	Programming language targeted at massive data processing and number crunching in the area of scientific computing. See the linked article for compiler support. All compiler suites mentioned for C/C++ also contain a Fortran compiler.
Fortran	Intel Compiler, GNU GCC, Portland Group Compilers, LLVM
Java	Object oriented, portable and powerful programming language for business applications, but also increasingly used in other areas.
Python	Interpreted, high-level, general-purpose programming language.
Julia	JIT compiled, multi-paradigm, general-purpose programming language (LLVM based)
Development Environments
Eclipse CDT	IDE (Integrated Development Environment) for Java/C/C++/Fortran
Intel Parallel Studio XE	Intel programming environment that comes in three flavors and is comprised of Intel Compilers: highly optimizing Fortran/C/C++ implementations (Composer Edition, Professional Edition, Cluster Edition) Intel Performance Libraries (MKL, TBB, IPP, DAAL): Math Kernel Library - MKL, Threading Building Blocks - TBB, Intel Performance Primitives - IPP, Data, Analytics Library - DAAL (Composer Edition, Professional Edition, Cluster Edition) Performance and Correctness/Memory management Analysis/debugging tools: Inspector, VTune Amplifier, Advisor (Professional Edition, Cluster Edition) Cluster software development: Intel MPI, Intel Trace Collector and Analyzer, Cluster Checking (Cluster Edition) All components are available on LRZ's HPC system. For licensing on the campus please consult the relevant document.
Programming Tools
Product	Short description
Doxygen	Semi-automatic generation of HTML or PDF Documentation from (possibly annotated) source code
Eclipse CDT	IDE (Integrated Development Environment) for C/C++/Fortran/Java
Forcheck	Tool for static analysis of Fortran code
PAPI	Performance Application Programming Interface: Evaluate hardware performance counters
Valgrind	Tool for correctness checking of programs (memory management, threading, etc.) that uses a virtual execution environment
Parallel Programming	Languages, Libraries and Tools for parallel programming
SVN (Subversion)	Software for source code versioning
git	Software for source code versioning
Debuggers (interactive and batch)
Product	Short description
gdb	GNU debugger
(x)pgdgb	PGI debugger
ddt	Allinea DDT (mainline debugging tool on LRZ HPC systems)
totalview	Rogue Wave Totalview (support expires in late 2015)
Scripting
Product	Short description
Rmpi	MPI extension for the statistical package R