Little’s law
Little’s law relates average items in a stable system to average arrival rate and average time spent there: L = λW.
How it works
The counted items and system boundary must match. For a memory path, a bandwidth-delay product estimates how many bytes need to remain in flight to sustain a rate. It is an average relationship, not a guarantee that any launch supplies enough independent requests.
Work through an example
An illustrative 1 TB/s path with 400 ns average latency needs about 400,000 bytes in flight: 10¹² bytes/s × 400 × 10⁻⁹ s. This does not directly specify a warp count.
Where the term can mislead
Do not mix a peak bandwidth with an unrelated latency measurement and call the result measured concurrency. Queue stability and consistent populations matter.
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.