Parallel Programming on Multi-Core CPUs using Pthreads, MPI, intelTBB & OpenMP for PEMG-2010 workshop, a National Level Technology workshop.
Heterogeneous programming on CPU using MPI and intel TBB. Developed a program to implement Vector-Matrix multiplication. It was added to PEMG-2010 hands-on session programs.
Multi-producer and Multi-consumer program with large number of insertion and deletion with "Thread Affinity" setting to study the 'Performance' of a program bound to multiple Cores.
Boyer Moore Algorithm, a fast string searching algorithm, implemented in OpenCL, a GPGPU open standard.
Assisted in organizing PEMG (Performance Enhancement on Multi-Core Processors & GPUs) at Electronic City, Bangaluru, from June 21-24, 2010.
Delivered talk on Performance & Tuning issues related to thread-synchronization-intensive Parallel-Programs. Demonstrated the Performance and Scalability of Multi-Producer/Multi-Consumer Program when bound to different number of cores.
Worked on various HPC (High Performance Computing) benchmarks: LINPACK, ATLAS, LAPACK BLAS, dhrystone, whetstone, STREAM, PARDISO, lmbench, DGEMM
Tuning and Performance on Multi-Core processors.
- nVidia CUDA