This is a project for the GPU COMPUTING course at the University of Milan, M.Sc. in Computer Science, A.Y. 2020/2021. The project I made consists of a comparison between a CPU implementation of the Ray Tracer and a GPU implementation made with CUDA. To read the report just click this link.
nvcc main.cu -o rt
nvprof --metrics achieved_occupancy,inst_executed,inst_fp_32,inst_fp_64,inst_integer,dram_read_throughput,dram_write_throughput,ipc,ipc_instance,sm_efficiency,warp_execution_efficiency ./rt
For other information about the comparison with the CPU version of the same scene, check the PDF paper.