Computational Resources

CATCO's computer resources include state of the art a symmetric multiprocessing and shared memory computer cluster. The resources of individual nodes are combined via an aggregated virtual machine that allows a single operating system instance to run above the discrete machines comprising the cluster.

Total:
Nodes 16
CPU's 32
Cores 128
Simultaneous Threads 256
Memory 768 GB
Hard Drive Capacity 16 TB
 
Per Node:
CPU's 2
Cores 8
Simultaneous Threads 16
Memory 48 GB
Hard Drive Capacity 1 TB
Network Interconnects 32 Gbit/s

Researchers of the CATCO group have access to the shared SMU research computing cluster (SMUHPC):

  • 163 batch worker nodes, connected with a gigabit ethernet network:
    • 107 are 8-core Intel Xeon 2.53 GHz nodes with 48 GB RAM each.
    • 56 are 12-core Intel Xeon 2.8 GHz nodes with 72 GB RAM each.
  • 48 MPI nodes, connected with an Infiniband high-speed network:
    • 16 are 8-core Intel Xeon 2.53 GHz nodes with 48 GB RAM each.
    • 32 are 12-core Intel Xeon 2.8 GHz nodes with 72 GB RAM each.
  • 2 high memory data analysis and shared-memory parallel nodes, each with 2 quad-core Intel Xeon 2.66 GHz processors, 144 GB RAM, and 3 TB of local disk space.
  • 1 GPU computing node, with 2 quad-core Intel Xeon 2.26 GHz processors, and 2 NVIDIA GTX 295 cards (each has 960 GPU cores).
  • One 320 TB parallel file system (Lustre) is attached to all nodes.
  • OS: Scientific Linux 5.5 (64 bit).
  • Usage: research.
  • Scheduler: Condor.
  • Software: Matlab, GNU compilers, PGI compilers, NAG compilers, hdf5, fftw, mpich2, mvapich2, python, R, ...

For more information on the SMUHPC cluster, see the SMU HPC Wiki. For access to this system, contact Justin Ross (ross [at] smu.edu).