000212688 001__ 212688
000212688 005__ 20181007165050.0
000212688 037__ $$aCONF
000212688 245__ $$aGPUfs: Integrating a File System with GPUs
000212688 269__ $$a2013
000212688 260__ $$c2013
000212688 336__ $$aConference Papers
000212688 520__ $$aAs GPU hardware becomes increasingly general-purpose, it is quickly outgrowing the traditional, constrained GPU-as-coprocessor programming model. To make GPUs easier to program and improve their integration with operating systems, we propose making the host’s file system directly accessible to GPU code. GPUfs provides a POSIX-like API for GPU programs, exploits GPU parallelism for efficiency, and optimizes GPU file access by extending the host CPU’s buffer cache into GPU memory. Our experiments, based on a set of real benchmarks adapted to use our file system, demonstrate the feasibility and benefits of the GPUfs approach. For example, a self-contained GPU program that searches for a set of strings throughout the Linux kernel source tree runs over seven times faster than on an eight-core CPU.
000212688 700__ $$aSilberstein, Mark
000212688 700__ $$aFord, Bryan
000212688 700__ $$aKeidar, Idit
000212688 700__ $$aWitchel, Emmett
000212688 7112_ $$a18th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS 2013)$$cHouston, TX, USA$$dMarch 16-20, 2013
000212688 8564_ $$uhttp://dedis.cs.yale.edu/2010/det/papers/asplos13-gpufs-abs$$zURL
000212688 8564_ $$s551317$$uhttps://infoscience.epfl.ch/record/212688/files/asplos13-gpufs.pdf$$yPublisher's version$$zPublisher's version
000212688 909C0 $$0252572$$pDEDIS$$xU13061
000212688 909CO $$ooai:infoscience.tind.io:212688$$pconf$$pIC$$qGLOBAL_SET
000212688 917Z8 $$x257875
000212688 917Z8 $$x148230
000212688 937__ $$aEPFL-CONF-212688
000212688 973__ $$aOTHER$$rREVIEWED$$sPUBLISHED
000212688 980__ $$aCONF