RMK001: loop nesting that might benefit from hybrid parallelization using multithreading and SIMD

warning

This check was deprecated in favor of PWR050, PWR051, PWR052, PWR053, and PWR054. This is because RMK001 combined multithreading and vectorization concerns into a single check. The highlighted checks split these cases into more focused and actionable rules.

Issue

Use multi-threading and SIMD hybrid parallelization to maximize the parallel performance of the nested loop.

Actions

Parallelize the outer loop using multi-threading and the inner loop using vectorization.

Relevance

Certain nested loops may benefit from using a hybrid parallelization approach where the outer loop is parallelized using multi-threading while using vectorization for the inner loop. In this way, the use of modern hardware is maximized by using as many CPU cores as possible through multi-threading and at the same time using the vectorization hardware available in each core.