An ECM-based energy-efficiency optimization approach for bandwidth-limited streaming kernels on recent Intel Xeon processors

摘要

We investigate an approach that uses low-level analysis and theexecution-cache-memory (ECM) performance model in combination with tuning ofhardware parameters to lower energy requirements of memory-bound applications.The ECM model is extended appropriately to deal with software optimizationssuch as non-temporal stores. Using incremental steps and the ECM model, weanalytically quantify the impact of various single-core optimizations andpinpoint microarchitectural improvements that are relevant to energyconsumption. Using a 2D Jacobi solver as example that can serve as a blueprintfor other memory-bound applications, we evaluate our approach on the four mostrecent Intel Xeon E5 processors (Sandy Bridge-EP, Ivy Bridge-EP, Haswell-EP,and Broadwell-EP). We find that chip energy consumption can be reduced in therange of 2.0-2.4$\times$ on the examined processors.