Multi-threaded adaptation of libcinder's ParticleSphereCPU sample. While, obviously the correct approach involves moving as much work to GPU as possible, I thought it might be fun to build a multi-threaded CPU implementation. Spawns a thread for each hardware core as reported by std::thread::hardware_concurrency() to update a subset of the particle state. WHile the threaded implementation is not non-blocking in a strict sense, it does not present any deadlocking writer/reader scenarios.
- Clone this repo into
{CINDER_PATH}/samples/_opengland openxcode/ParticleSphereCPU_threaded.xcodeproj - Select the
ParticleSphereCPU_threadedbuild scheme and run