CUDA & Parallel Processing
Presented and led discussions regarding foundations and paradigms of GPU computing, provisioning an environment for CUDA development, and introductory material for developing GPGPU applications.
Aided in business development and planning across multiple projects to include system integration, deployment preparation, and live demonstrations for customers across multiple agencies.
Integral member of a cluster application with focus on 'Big Data' analysis and analytic fusion; lead teams in data ingestion strategy and design, analytic development and production integration, as well as scalable architectures.
System administrator over a 20-node cluster running Hadoop, Zookeeper, and Accumulo to perform network defense capabilities while maintaining federal government Red Hat security standards and practices.
Master of Science
Developed a complete distributed environment to compute the PageRank algorithm leveraging OpenMPI, Qsub / Torque, and Eucalyptus as well as a graphical monitoring component through Java.
Ran extensive benchmarking and testing analyses running against the PageRank algorithm on physical nodes compared to Eucalyptus-hosted virtual instances.
Became familiar with core distributed computing concepts such as timing (Lamport timestamps), synchronization (distributed locking), and fault tolerance (Byzantine General’s problem).
Hypervisor Security Survey
Investigated three major Type 1 Hypervisors with emphasis on their security capabilities and concerns including Xen, VMware ESX(i), and Microsoft Hyper-V.