Partners Research Computing

Partners High Performance Computing Cluster

Partners applications via Mac/PC

Clinical & research applications

DFCI bioinformatics computer

PHS Research Computing cluster

Bioinformatics news

Data Storage & Backup

Sharing files & collaboration

HIPAA, ePHI and research (internal)

RPDR

HPCGG

Biomedical Engineering Model Shop

Harvard's GForge Implementation

Institutional research distribution lists

pHPC account registration	pHPC user guide	pHPC services	pHPC web protal

Setup your Intel Compiler enviroment on hpres cluster

Intro: The HP XC-4000 cluster by default provides GNU compiler (gcc3.4) in your working environment. We also have installed Intel ICC C/C++/FORTRAN compiler in the cluster. The performance of your program/application can be averagely improved 50% to 10 times higher comparing to GNU compiler.
More instructions can be found online at HPC Web Portal (in PHS network)

More examples can be found at /source/intel/cce/10.1.008/samples in the system.

Step 1: Please read "how to setup your working environment" and do the following:

check your icc path.

[testy@n137 ~]$ which icc
/source/intel/cce/10.1.008/bin/icc

This indicates your environment has been setup correctly and you can begin to work and submit your job.

Step 5. Compiling your program

example int_sin.c can be downloaded.

[jxu@n137 optimize]$ icc -o int_sin int_sin.c -lm
int_sin.c(92): (col. 6) remark: LOOP WAS VECTORIZED.

Run your program
( Attention: The followiing run on login node is just about testing the compiler. If you have real job need to be dispatched to the cluster, please refer to other pages about "how to submit job" or "how to run massive jobs" )
[testy@n137 ~]$ ./sin Number of | Computed Integral | Interior Points | | ------------------------------------- 4 | 3.141593e+00 | ------------------------------------- 8 | 3.792238e+00 | ------------------------------------- 16 | 3.948463e+00 | ------------------------------------- 32 | 3.987141e+00 | ------------------------------------- 64 | 3.996787e+00 | ------------------------------------- 128 | 3.999197e+00 | ------------------------------------- 256 | 3.999799e+00 | ------------------------------------- 512 | 3.999950e+00 | ------------------------------------- 1024 | 3.999987e+00 | ------------------------------------- 2048 | 3.999997e+00 | ------------------------------------- 4096 | 3.999999e+00 | ------------------------------------- 8192 | 4.000000e+00 | ------------------------------------- 16384 | 4.000000e+00 | ------------------------------------- 32768 | 4.000000e+00 | ------------------------------------- 65536 | 4.000000e+00 | ------------------------------------- 131072 | 4.000000e+00 | ------------------------------------- 262144 | 4.000000e+00 | ------------------------------------- 524288 | 4.000000e+00 | ------------------------------------- 1048576 | 4.000000e+00 | ------------------------------------- 2097152 | 4.000000e+00 | ------------------------------------- 4194304 | 4.000000e+00 | ------------------------------------- 8388608 | 4.000000e+00 | ------------------------------------- 16777216 | 4.000000e+00 | ------------------------------------- 33554432 | 4.000000e+00 | ------------------------------------- 67108864 | 4.000000e+00 | Application Clocks = 1.430000e+06
If you have above results, it indicates you are ready to use intel compiler .

In most cases, particularly for situations with lots loops, simulations like Monte Carlo, Intel Compiler can dramatically improve the performance of your application. more examples can be found in /source/intel/cce/10.1.008/samples