[FEA]: Add SM90 to architecture list for CI

[jrhemstad]

Is this a duplicate?

• I confirmed there appear to be no duplicate issues for this request and that I agree to the Code of Conduct

Area

Infrastructure

Is your feature request related to a problem? Please describe.

As a user and maintainer of CCCL, I want to be sure that CCCL works when compiled and run for SM90 (Hopper). However, CCCL's CI does not currently compile (or run) for SM90.

There are a number of things that complicate simply adding SM90 to the architecture list:

• Compiling RDC tests in Thrust/CUB break for SM90 today: [BUG]: Thrust/CUB should check that their CUDA_ARCHITECTURES_RDC list is non-empty before adding RDC targets. #219
• The CUDA architecture list is hard-coded in the CMake preset and there's no longer a way to customize the GPU architectures in the CI matrix. This functionality used to exist, but was removed with the addition the CMake preset.
• Simply adding SM90 to the build list will still not exercise compiling (or running) SM90 specific tests in libcudacxx due to the current design of libcudacxx's testing infrastructure

Describe the solution you'd like

Tasks

• Enable specifying CUDA architecture per CI job
• [FEA]: Remove CDP (RDC) architecture filtering logic from Thrust/CUB tests #1137
• [BUG]: Thrust/CUB should check that their CUDA_ARCHITECTURES_RDC list is non-empty before adding RDC targets. #219
• Redesign libcudacxx architecture specific testing #1084

Describe alternatives you've considered

No response

Additional context

No response