Pipeline utilization
Pipeline utilization describes activity or throughput on a particular execution pipeline, under the profiler’s specified normalization.
How it works
Different pipelines handle different instruction classes. A metric may express issued work relative to a sustained peak, or report a cycle-based activity measure. Read the exact identifier, unit, aggregation and active-versus-elapsed suffix before interpreting its percentage.
Work through an example
A busy tensor pipeline and a lightly used scalar pipeline can be expected in a matrix-heavy kernel. Averaging the two without their definitions would produce a misleading overall efficiency score.
Where the term can mislead
High pipeline utilization does not prove the work was necessary. Redundant operations can keep a resource busy.
Connect the idea
Take it into practice
Continue with the existing concept, apply it to an interview problem, or follow a course lesson. Access requirements are shown before you open a page.
New to GPUs? Start with the free execution-model concept.
Sources and scope
Examples here are teaching calculations, not measured GPU performance. Consult the linked documentation for the exact device, instruction and tool version. Source review: 25 October 2026.