GPU Gems 3

536 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Fast geometric learning with symbolic matrices (2020, methodology): “These results highlight the dichotomy between contiguous (“bruteforce”) and scattered (“sparse”) memory accesses, which are optimized separately by GPUs and are best suited to different types of computations.”
Kernel Operations on the GPU, with Autodiff, without Memory Overflows (2020, methodology): “KeOps leverages eﬃcient C++ schemes from the graphics literature (Nguyen, 2007, Chapter 31) to streamline the use of registers and reach optimal performance in this setting.”
Computational Physics on Graphics Processing Units (2012, background): “They were proved to give a good acceleration in solving the N-body problem [13,30,31].”
Monte Carlo simulation of photon migration in 3D turbid media accelerated by graphics processing units (2009, methodology): “Fortunately, an efficient parallel version of MT known as SFMT (SIMD-oriented Fast MT) [25–27], was made available by Mills [28] using a shared-memory implementation.”