Manuals for SuperMUC
Table of contents
Access to Manuals
Please note that material provided here underlies the copyright of its respective producer/owner; it may usually not be republished and/or redistributed. Hence, please only use this material for your own reference.
Some of the documentation on this page is password protected. Please log in to the HPC system and type
to obtain the user name and password required for validation.
Optimization and Tuning
- High Performance Computing Central is a joint IBM/Customer accessible and editable forum to provide improved HPC technical communications: HTML
Intel Compiler, Libraries and Tools
Intel Optimization and Tuning
- Quick-Reference Guide to Optimization with Intel® Compilers
- Compiling for AVX
- Guide to Auto-Vectorization
- A Guide to Vectorization with Intel® C++ Compilers
- Requirements for Vectorizable Loops
- Developing Multithreaded Applications: A Platform Consistent Approach
- Consistency of Floating-Point Results Using the Intel® Compiler
Intel 64 and IA-32 Architectures Software Developer's Manuals
SuperMUC Best Practice Guide
SuperMUC Best Practice Guide written within PRACE-2IP
Other Best Practice Guides written within the PRACE project can be found under PRACE Best Practice Guides.
for the SuperMUC workshop "Introduction to SuperMUC"
Thanks are due to H. Bast and G. Zitzlsberger (Intel), A. Bömelburg, H. Holthoff, F. Merz and C. Pospiech (IBM) for preparing the course materials and making them available.
Intel development software talks
- Overview of Intel Software Development Products
- Intel Architecture Update (Sandy Bridge and MIC)
- Intel Compilers for Linux
- Some advanced features of the compilers
- Fortran-specific (ifort) features and recommendations on extensions (R. Bader, LRZ)
- Intel debugger for Linux
- Instruction Set Extensions for the x86 architecture (SSE, AVX)
- Introduction to Vectorization
- Data and Task parallelism with Cilk Plus
- Non-Uniform Memory Access (NUMA)
- Intel Threading Building Blocks and an update on TBB version 4 features.
- Intel Math Kernel Library
- Intel Performance Primitives
- Correctness checking with Intel Inspector XE
- Investigating parallelization potential with Intel Advisor XE
- Finding performance bottlenecks with Vtune Amplifier XE
- Application optimization using Performance Monitoring Events
- Intel Cluster Tools, more specific information on Trace Analyzer
- 21 tips for performance tuning with the Intel MPI library
- Some Experimental Tools from Intel
- Programming for the Intel Many Integrated Core (MIC) Architecture
- Case Study: Using Intel Parallel Studio to optimize Black Scholes calculation. Sample code also provided in ZIP format.
IBM HPC software talks
These talks are presently only partially available.
- Introduction to the IBM Parallel Environment
Christoph Pospiech's talks are based on the following materials: