How work runs · free reference

Thread block

A thread block is a group of CUDA threads that can share block-scoped memory and synchronize together. PTX commonly calls it a cooperative thread array, or CTA.

BlockCTACooperative thread array

How it works

An ordinary block is assigned to one SM. Its threads share a lifetime for block resources, so the block’s register and shared-memory requirements affect how many blocks can reside together. A block barrier coordinates participating threads before later accesses.

A launch is software work. An SM is hardware.

01 · LaunchGridFour blocks in this example.02 · DivideBlock256 threads assigned to one SM.03 · GroupWarp32 threads; eight warps per block.04 · ExecuteSMResident warps share its execution resources.

A launch for 1,000 elements with 256 threads per block creates 1,024 logical threads. The final 24 skip the array access. Blocks can run in successive waves; this diagram does not imply four physical SMs or simultaneous execution.

Work through an example

If a block loads a tile into shared memory, all required writes must finish before threads read one another’s values. Place synchronization on a control-flow path all required threads can reach.

Where the term can mislead

A block barrier does not synchronize the entire grid. Separate kernel launches or supported cooperative mechanisms are needed for wider coordination.

Connect the idea

Grid →Shared memory →Occupancy →

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.

ConceptCUDA Programming ModelFree ConceptShared Memory and Bank ConflictsPremium

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.