From Lustre Wiki
Jump to navigation Jump to search

Senior HPC Systems Architect (Firstspot/Naval Postgraduate School) • Design and implement HPC clusters using dense computing systems (64 cores/node), Mellanox infiniband switches, Force10 network switches, nVidia Tesla GPU cards, DDN lustre based disk storage, processor dense AMD computing nodes • Implement and maintain software stack for HPC systems to include PGI/Intel compilers, OpenMPI, MVAPICH2, MPICH2 transports • Design and Implement monitoring processes based on nagios, ganglia, snmp and web-based CGI scripts. • Manage cluster resource manager/queuing system based upon Torque/MOAB. • Manage cluster nodes via PXEBOOT using industry standard tools (Rocks, Warewulf, xCAT, Bright) • Create RPMs and repositories to manage software packages