arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1909.07190cs.PLcs.DC
收藏

GPU上基于模型的翘曲重叠拼接图像处理程序

Model-Based Warp Overlapped Tiling for Image Processing Programs on GPUs

Abhinav Jangda, Arjun Guha

摘要:

在GPU上执行图像处理流水线的领域特定语言(如Halide和Forma)通过以下方式进行操作:1)将图像划分为重叠的区块,以及2)融合循环以提高内存局部性。然而,当前的方法具有局限性:1)它们需要线程内块同步,这具有不小的成本,2)它们必须在需要更多重叠计算的小区块或增加共享存储器访问(并降低占用率)的大区块之间进行选择,以及3)它们的自动调度算法使用简化的GPU模型,这可能导致低效的全局存储器访问。我们提出了一种在GPU上执行图像处理流水线的新方法,该方法解决了这些限制,如下所述。1)我们融合循环以形成重叠的瓦片,这些瓦片适合单个经线,这允许我们使用轻量级经线同步。2)我们引入混合平铺,它存储在线程本地寄存器和共享内存的组合重叠的区域。因此,混合平铺要么通过减少共享内存使用来增加占用率,要么使用更大的平铺来减少重叠计算。3)我们提出了一个自动循环融合算法,考虑了影响GPU内核性能的几个因素。我们实现了这些技术在PolyMage-GPU,这是一个新的GPU后端PolyMage。我们的方法生成的代码比Halide的手动时间表更快:在NVIDIA GTX 1080Ti上快1.65倍,在NVIDIA Tesla V100上快1.33倍。

英文摘要:

Domain-specific languages that execute image processing pipelineson GPUs, such as Halide and Forma, operate by 1) dividing the image into overlapped tiles, and 2) fusing loops to improve memory locality. However, current approaches have limitations: 1) they require intra thread block synchronization, which has a non-trivial cost, 2) they must choose between small tiles that require more overlapped computations or large tiles that increase shared memory access (and lowers occupancy), and 3) their autoscheduling algorithms use simplified GPU models that can result in inefficient global memory accesses. We present a new approach for executing image processing pipelines on GPUs that addresses these limitations as follows. 1) We fuse loops to form overlapped tiles that fit in a single warp, which allows us to use lightweight warp synchronization. 2) We introduce hybrid tiling, which stores overlapped regions in a combination of thread-local registers and shared memory. Thus hybrid tiling either increases occupancy by decreasing shared memory usage or decreases overlapping computations using larger tiles. 3) We present an automatic loop fusion algorithm that considers several factors that affect the performance of GPU kernels. We implement these techniques in PolyMage-GPU, which is a new GPU backend for PolyMage. Our approach produces code that is faster than Halide's manual schedules: 1.65x faster on an NVIDIA GTX 1080Ti and 1.33 faster on an NVIDIA Tesla V100.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑