Memory bandwidth
Memory bandwidth is the rate at which bytes cross a particular memory interface. A specification gives a peak; a benchmark or profiler gives a rate under stated conditions.
How it works
Reads and writes, request size, access pattern, caching and concurrency affect achieved bandwidth. Application-effective bandwidth can count only useful bytes, while hardware counters may count additional transferred bytes.
Work through an example
If a measured interval moves 600 MB of counted traffic in 1 ms, the rate is 600 GB/s in decimal units. State whether those bytes are useful payload, cache sectors or physical DRAM traffic.
Where the term can mislead
Bandwidth is not capacity. An 80 GB allocation limit and a 3 TB/s interface rate describe different quantities.
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.