Haleh Najafzadeh

Learn More
Microprocessors typically have software readable counters for events such as instruction executions, cycles, instruction stalls, and cache misses. Besides their usefulness to report overall performance metrics, these counters reveal details about dynamic process behavior and hardware affects of compiler optimizations. Our research develops and evaluates, in(More)
Ignoring monitoring overhead and not validating measurements are two common mistakes in benchmarking. We extend, apply and evaluate for a practical FFT library our own methodologies for overhead compensation and over-all validation. The overhead and error sources we address include the source instrumentation and separate activities unrelated to the(More)
A General Partitioning Algorithm can provide a powerful way of computing local and global addressing for diierent distributions on diierent processor topologies. Therefore, it can be used to design an eecient algorithm for run-time (re)distributions in array based languages such as High Performance Fortran(HPF). EEcient redistributions are essential in many(More)
  • 1