Formosa II HPC Cluster

Detail

Formosa II HPC Cluster is a 64bits high-performance Beowulf cluster located within Southern Business Unit of the National Center of High-Performance Computing (NCHC). 'SIRAYA' is the nickname for Formosa II cluster, and it means that Siraya were an indigeneous people of Taiwan. They lived in the southwest part (Tainan) of the island. The Siraya cluster was designed and constructed by the 'HPC Cluster Group' at NCHC for computational science applications. It consists of 80 dual AMD Opteron DualCore PCs (i.e. 320 2.2GHz CPU-Cores) connected on a private subnet with 1000 Mbits/s Gigabit Ethernet. An additional 2 nodes are used as front ends to interface with cluster, to host the user file systems.

  Hardware 
     Computing Nodes -80 computing nodes of IBM eServer 326. Each computing node consists of:

1. CPU – AMD Opteron 275 DualCore 2.2GHz x 2
2. Main Board – IBM eSeries e326
3. Memory – 4GB DDR400 Registered/ECC SDRAM
4. Hard Disk – SCSI 36GB 

     Storage Nodes - 2 storage nodes.

1. CPU – AMD Opteron 248 2.2GHz x 2
2. Main Board – Supermicro X5DPE-G2
3. Memory – 8GB DDR400 Registered/ECC SDRAM
4. Hard Disk – ATA-100 SATA 400GB x 16
5. SATA card – Serial ATA Raid card x 2
6. Network card – Intel PRO/1000 MT Server Adapter x 2 
    

  Network
     Gigabit switch – Nortel Baystack5510 -48 ports 10/100/1000Mbps switch x 6

  Operating System
     Fedora Core 4 X86_64: Linux kernel 2.6.14 SMP

  Compilers 

     C/C++ Compiler

1. GNU GCC Collection 4.02 X86_64 (gcc,g++)
2. Intel C/C++9.0 EM64T (icc)
3. PGI C/C++6.0 X86_64 (pgc,pgcc)
4. PathScale 2.53 X86_64 (pathcc,pathCC) 

      Fortran

1. GNU GCC Collection 4.02 X86_64 (gfortran)
2. Intel Fortran 9.0 EM64T (ifc)
3. PGI PGF77 6.0 X86_64 (pgf77,pgf90)
4. PathScale Fortran 2.53 X86_64 (pathf90,pathf95)

  Libraries
     MPICH 1.2.7pl

1. MPICH compiled with GCC Compiler
2. MPICH compiled with Intel Compiler
3. MPICH compiled with PGI Compiler
4. MPICH compiled with PathScale Compiler

     LAM/MPI 7.11

1. LAM/MPI compiled with GCC Compiler
2. LAM/MPI compiled with Intel Compiler
3. LAM/MPI compiled with PGI Compiler
4. LAM/MPI compiled with PathScale Compiler

  Queuing System:
     Torque 2.0.0

  Scheduler:
     Maui 3.2.5

  Chemical Applications

1. ADF/BAND
2. CRYSTAL
3. AMBER 8
4. Gaussian 03/GaussView
5. DL_POLY
6. GROMACS
7. NWCHEM
8. Molpro 2002.6
9. CPMD
10. CHARMM
11. DMOl3
12. DISCOVERY
13. VASP
14. babel (Openbabel)
15. VMD
16. gabedit
17. molekel
18. molden  

   For related detail information, please refer to http://siraya.sro.nchc.org.tw/