PTX - what is a CTA?

Question

I am studying PTX, and I do not understand how CTA (Computed Stream Array) differs from a CUDA block.

They are the same? It seems to me that at the moment (I'm only at the beginning of the PTX document) they are all the same

Answer 1

Yes, the PTX stream computing array conceptually and functionally matches a block in CUDA or a workgroup in OpenCL.

Answer 2

the thread hierarchy section of the CUDA PTX ISA document explains that essentially a CTA means a CUDA block .

Also note that this is actually not a “computational array of threads,” but rather a “cooperative array of threads” (!).