The Itanium2 processors can be a bit more difficult to get your code to run on than a Xeon or Pentium, but the reward for this extra work can be very fast execution!
A good strategy is to first use the Gnu compilers, and compile with debug (the -g flag). Once the code is running, move to the Intel compilers with -g, and once that works, finally start adding optimization flags.
We have both the GNU 2.96 and Intel 7.1 compilers - the Intel compilers generally give the best performance. We also have the highly optimized Intel MKL math libraries - you can see significant performance improvements by linking to these libraries (see the special libraries section)
The GNU compilers are accessed by simply "g77" for FORTRAN, or "gcc" for C.
The Intel compilers are "efc" and "ecc". You do need to have /usr/local/intel/compiler70/ia64/bin in your path. This is usually done in your ".cshrc" file.
Compiler Flags
| Compiler Flag | Description |
|---|---|
| -g | For debugging |
| -O2 | Standard optimization, -O3 provides more aggressive optimzation |
| -tpp2 | Option for Itanium2 processor |
| -openmp and -parallel | Parallel processing directives (see parallel/MPI coding tab for more info) |
| -ipo | Interproceedure optimizations and inlining |
| -W0 | Turn off warning messages |