NERSCPowering Scientific Discovery Since 1974

Performance and Optimization

Compiler Comparisons

Comparison of different compilers with different options on several benchmarks. Read More »

Using OpenMP Effectively

Performance implications and case studies of codes combining MPI and OpenMP Read More »

Reordering MPI Ranks

Reordering MPI ranks can result in improved application performance depending on the communication patterns of the application. Read More »

Application Performance Variability on Hopper

How an application is placed across Hopper's roughly 6300 compute nodes can affect its performance. See a study of application runtimes vs node placement. Read More »

Hopper Performance Monitoring

Benchmarking performance of scientific applications on Hopper Read More »

Hopper:Improving I/O performance to GSCRATCH and PROJECT

This page describes how to get the best performance when using GPFS file systems such as GSCRATCH and PROJECT on Hopper. Read More »