111 batch running features gpu runner

[jmaack24]

Implement/do the following for the OptixRunner:

• Improve performance of SolTraceSystem::get_buffer_results
• Add user specified batch size controlling number of rays traced per optixLaunch
• Add logic to estimate batch size to fit on available device memory
• Checks for GPU memory limit (i.e. Check available device memory, calculate required buffer size, batch accordingly)
• Checks for rays that go past max depth (may just want to warn user)
• Add user input for max depth

[Copilot]

Pull request overview

Copilot reviewed 38 out of 38 changed files in this pull request and generated 4 comments.

[Copilot]

Pull request overview

Copilot reviewed 43 out of 43 changed files in this pull request and generated 7 comments.

[jmaack24]

To avoid duplicating work, I pulled in changes made on #142. Some of these features (generally, intersection functions for new geometries), are incomplete. These will be completed by #142.

[taylorbrown75]

Tests pass on my windows machine. Only concern is whether we mark rays as EXIT on the gpu runner.

[taylorbrown75]

Is HIT_EXIT ever assigned in the GPU runner? The miss function in materials.cu does not set that value currently. There could be cases where rays exit the scene on the last available depth slot, but because it is not absorbed or marked as exit, it will be added to passed max depth counter.

[jmaack24]

SolTrace/coretrace/simulation_runner/optix_runner/OptixCSP/src/shaders/materials.cu

Line 291 in dac334b

atomicAdd(reinterpret_cast<unsigned long long*>(params.d_depth_exceeded_count), 1ULL);

The atomic counter for the max depth exceeded is incremented in the materials.cu (above). If I understand the flow correctly, it should only get there if there is an intersection with an element. In which case that is not registered because it exceeds max depth. So only non HIT_EXIT rays should get counted.

Currently, we do not assign HIT_EXIT in the optix runner.

[taylorbrown75]

This makes sense to me, thank you for the clarification!