Brook for GPUs: Stream Computing on Graphics Hardware

By Ian Buck, Theresa Foley, Daniel Reiter Horn, Jeremy Sugerman, Kayvon Fatahalian, M. Houston, P. Hanrahan. ACM Transactions on Graphics, 2004

1,430 citations recorded by the source corpus.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/cdf7b5c807ced81147b50de2f366a3fe2ba58c70

How researchers see this paper

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

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Benchmarking modern multiprocessors (2011, background): “Stream programming is a parallelization approach that decomposes a program into a parallel network of specialized kernels which are then mapped to processing elements [18, 29, 46].”
Dandelion: a compiler and runtime for heterogeneous systems (2013, methodology): “GPU frameworks such as CUDA [79], OpenCL [63], and others [88, 17, 71, 22, 48, 94] provide rich front-end programming models, specific to the underlying parallel GPU architecture.”
Dynamic Warp Formation and Scheduling for Efficient GPU Control Flow (2007, methodology): “Until recently, the dominant approach has been to extract more instruction level parallelism from a single thread through increasingly complex scheduling logic and larger caches.”
Nodal discontinuous Galerkin methods on graphics processors (2009, neutral): “In the last few years, high-level abstractions such as Brook and Brook for GPUs [2] have enabled more and more complex computations on streaming hardware.”
Mars: A MapReduce Framework on graphics processors (2008, methodology): “With the MapReduce framework, the user does not require the knowledge of a speci.c programming model such as the streaming programming model in Brook.”
GPU Cluster for High Performance Computing (2004, background): “Recently, the use of the GPU to accelerate non-graphics computation has drawn much attention [6, 16, 3, 29, 10, 28].”
Towards dense linear algebra for hybrid GPU accelerated manycore systems (2009, background): “Examples are the Brook [3] and Sh [15] languages along with their commercialized generalizations correspondingly in PeakStream and RapidMind.”
Merge: a programming model for heterogeneous multi-core systems (2008, methodology): “Direct approaches, such as the DSVMs cited previously or streaming languages (4; 30; 15; 20; 24), map the application directly to the accelerator ISA or API, while library-based approaches, such as the Merge framework, Sequoia, and MATCH (23), map the application to a library of function-intrinsics ”
TeraFLOP computing on a desktop PC with GPUs for 3D CFD (2008, neutral): “The use of the GPU to accelerate non-graphics computations has drawn much attention (Bolz et al. 2003, Kru¨ger and Westermann 2003, Buck et al. 2004).”
RPU: a programmable ray processing unit for realtime ray tracing (2005, background): “For instance, it allows for increasing the computational density [Buck et al. 2004] by reducing the amount of memory required to represent complex surfaces while increasing the amount of data parallel computation.”