Test01 - GPU Path Tracing - Starting Out

Channel:
Subscribers:
622
Published on ● Video Link: https://www.youtube.com/watch?v=7gVQd4V-Q7o



Duration: 2:15
7,122 views
137


My first attempt at porting my Monte Carlo-based path tracer over to CUDA to take advantage of the GPU compute performance.

The path tracer is 'brute force' right now and supports only simple primitives (including simple triangle meshes). Only a few basic material types are supported. An interactive camera allows the user to move around the 3D world in real time.

The rendered image converges over time as more samples are accumulated, and averaged. Movement in the scene causes the accumulation buffer to reset as the viewpoint has changed.

Written in C++ (Visual Studio 2015) and running on Windows 10 Pro (64-bit).

GPU - nVidia Geforce GTX 1080 Ti (EVGA FTW3)
CPU - Intel i7-6900K







Tags:
gpu
c++
programming
nvidia
path tracing
path tracer
raytracing
ray tracing
monte carlo
real time
3d
graphics
global illumination