NVIDIA Announces the Turing GPU Architecture

NVIDIA announced the Turing GPU architecture in August 2018, introducing dedicated hardware for real-time ray tracing and expanding Tensor Cores beyond the data-center-focused Volta generation.

RT Cores accelerated operations such as ray–triangle intersection tests and bounding-volume hierarchy traversal, two expensive parts of ray tracing. Tensor Cores handled matrix operations used by neural networks. NVIDIA combined both ideas in RTX features such as denoising and later DLSS, where machine learning helped reconstruct higher-quality images.

Turing changed the graphics roadmap. Rasterization remained dominant, but GPU design now included fixed-function hardware for physically based lighting and AI inference alongside traditional shader cores. Modern gaming GPUs have continued down that heterogeneous path.