Boost Performance and Energy Efficiency on HPC Workloads with Intel® Xeon Phi™ Processors

ID 672343
Updated 6/15/2016
Version Latest
Public

author-image

By

Testing on LAMMPS demonstrates the potential for demanding applications.

Modernizing code with increased parallelism not only allows software to take advantage of the full breadth of Intel® architecture hardware capabilities—it can also increase power efficiency. Optimizing code on the latest Intel® Xeon® E5 v4 and Intel® Xeon Phi™ processors helps transform manycore processing power into faster time to insight.

An evaluation of the performance of LAMMPS on the new Intel Xeon E5 v4 and Intel Xeon Phi processors found more than 7x increased simulations rates and more than 9x the CPU+DRAM power efficiency when compared to unoptimized code on previous generation processors.