Instant neural graphics primitives with a multiresolution hash encoding

TLDR

Neural graphics primitives, parameterized by fully connected neural networks, can be costly to train and evaluate. The authors aim to reduce this cost by introducing a versatile input encoding that allows a smaller network without sacrificing quality. They augment the small network with a multiresolution hash table of trainable feature vectors, disambiguating hash collisions and enabling a simple, GPU‑parallelizable architecture implemented with fully‑fused CUDA kernels to minimize bandwidth and compute waste. This approach yields several‑order‑of‑magnitude speedups, enabling high‑quality neural graphics primitives to be trained in seconds and rendered in tens of milliseconds at 1920×1080 resolution.

Abstract

Neural graphics primitives, parameterized by fully connected neural networks, can be costly to train and evaluate. We reduce this cost with a versatile new input encoding that permits the use of a smaller network without sacrificing quality, thus significantly reducing the number of floating point and memory access operations: a small neural network is augmented by a multiresolution hash table of trainable feature vectors whose values are optimized through stochastic gradient descent. The multiresolution structure allows the network to disambiguate hash collisions, making for a simple architecture that is trivial to parallelize on modern GPUs. We leverage this parallelism by implementing the whole system using fully-fused CUDA kernels with a focus on minimizing wasted bandwidth and compute operations. We achieve a combined speedup of several orders of magnitude, enabling training of high-quality neural graphics primitives in a matter of seconds, and rendering in tens of milliseconds at a resolution of ${1920\!\times\!1080}$.