CuTe
CuTe is a layout and tensor abstraction used in CUTLASS to describe how multidimensional coordinates map to data and execution resources.
How it works
A layout combines shape and stride information. Composing layouts helps express tiling and the distribution of values across threads and memory. The distinction between logical coordinates and physical addresses is central to understanding the generated accesses.
Work through an example
Two arrays can have the same logical shape while placing consecutive values at different address strides. Their layouts can therefore lead to different coalescing or shared-memory bank behavior.
Where the term can mislead
CuTe’s mathematical layout operations do not erase hardware constraints. Check the resulting addresses and instruction requirements.
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.