Principal Hpc Engineer
CurrentProviding support and guidance for scientific research on HPC resources. Maintaining 7000+ core Linux compute cluster as well as a gpu compute resource using IBM Spectrum LSF and historically Sun/Univa Grid Engine. Optimizing for high throughput compute jobs, averaging 130,000+ jobs run daily. Managing high performance storage including DDN GridScaler and Seagate ClusterStor based on GPFS. Supporting high volume data storage and stewardship in a multiprotocol environment on Isilon and Qumulo storage systems. Aggregating access to multiple file systems via Avere FXT edge filers for multiprotocol connectivity.Working with software developers as well as scientific researchers to optimize their processes to leverage our compute and storage resources to improve productivity, improve time to insight and maximize the hardware investment. Provide HPC scheduler domain expertise to enable integration of Spark into the general purpose compute cluster enabling the use of the Thunder library developed by the Freeman Lab.