Compute capability
Compute capability identifies a set of NVIDIA GPU hardware features and limits. Compilers use targets such as sm_90 to generate suitable native device code.
How it works
The toolkit version, driver version and compute capability describe different layers. An application needs code the device can execute, a toolchain that can build it, and a driver able to load that code. Some architecture-specific target suffixes impose narrower compatibility rules.
Work through an example
A package can contain several cubins plus PTX fallback. On a particular GPU, the driver chooses a compatible image or compiles supported PTX when that path is available.
Where the term can mislead
A higher-looking version string does not prove compatibility. Check the documented binary, PTX and architecture-specific target rules for the exact artifact.
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.