Pierre Moreau

1. and 3. will define the same distance, as the film is located in the focal plane IIRC. There are two focal planes located on each side of the lens...

In CMake 3.18, [`CMAKE_CUDA_ARCHITECTURES`](https://cmake.org/cmake/help/latest/variable/CMAKE_CUDA_ARCHITECTURES.html) was added which gives a way to the user to specify for which architectures to compile (and CMake will automatically forward those values to any CUDA...

From what I understand from the doc, specifying `CMAKE_CUDA_ARCHITECTURES="70-real;72-real"` will have CMake automatically forward those architectures to nvcc and build for those (I am guessing using its “fat binary” functionality)....

How would the unique naming work? Would `src/pbrt/gpu/aggregrate.cpp` need to to contain something like the following? ```cpp extern "C" { extern const unsigned char PBRT_EMBEDDED_PTX_SM30[]; extern const unsigned char PBRT_EMBEDDED_PTX_SM50[];...

Maybe trying adding that property to `pbrt_exe` instead, as that is the target containing the entry point and doing the argument parsing. ``` if (CMAKE_CXX_COMPILER_ID STREQUAL "MSVC") # Enable wildcard...

> (Among other things, the headers don't seem to appearing even though there's a source_group directive that should be putting them there.) Only the files that are added to the...

The 7.3s with the 3090 was on Linux I'm guessing? Just tested with 06000be on a 3080 on Windows for the crown scene, and it took ~1138s. The stats for...

I can confirm that this improves performance significantly, going from 1138s to 259s on my 3080 when running without `--spp 64`, and down to ~32s with `--spp 64` on the...

Regarding Windows (will reboot later to try it out on Linux), I am getting some variation but roughly in line with what you have as well; the following are all...

Here are the results from the same system, also running CUDA 11.2 and OptiX 7.2, but on Linux (and it rendered in 7-8s): ``` Wavefront Kernel Profile: Generate Camera rays...