The Best of Modern Code | April

 MPI-3 Shared Memory in Intel® Xeon Phi™ Processors

Using MPI-3 Shared Memory in Intel® Xeon Phi™ Processors

Introducing the MPI-3 shared memory feature that allows multiple MPI processes to allocate and have access to the shared memory in a compute node. Use the supplied sample code, which uses IMP-3 shared memory APIs.

Software Defined Visualization

Software-Defined Visualization Using the Intel® Xeon Phi™ Processor

To help advance research in the medical field, we show how the Intel® Xeon Phi™ processor powers 3D visualization of a tumor in the human brain.


Recipe: Building and Running NEMO* on Intel® Xeon Phi™ Processors

The steps in this recipe demonstrate the performance advantages of using the Intel® Xeon Phi™ processor 7250 to run the NEMO* (Nucleus for European Modeling of the Ocean) numerical solutions framework.

Good Syntax?

The Unfairness of Good Syntax

Good syntax does not solve programming problems; programming remains difficult. However, bad syntax can really get in the way. Find out what you can do to remove bad syntax roadblocks.

Performance Portability

Rainbows, Unicorns, and Performance Portability

How much performance would you be willing to give up by replacing two optimal applications by a single one? Learn more about parallelization for both CPU and GPUs.

For more complete information about compiler optimizations, see our Optimization Notice.