Performance Analysis on Energy E cient High-Performance Architectures

With the shift in high-performance computing (HPC) towards energy e cient hardware architectures such as accelerators (NVIDIA GPUs) and embedded systems (ARM processors), arose the need to adapt existing performance analysis tools to these new systems. We present EZTrace a performance analysis framework for parallel applications. EZTrace relies on several core components, in particular on a mechanism for instrumenting functions, a lightweight tool for recording events, and a generic interface for writing traces. To support EZTrace on energy e cient HPC systems, we developed a CUDA module and ported EZTrace to ARM processors. The evaluation on a suite of the standard computation kernels show that EZTrace allows to analyze HPC applications executing on such systems with the low performance overhead.