Pagoda
Authors:
Open Access
- 26 January 2017
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 221-234
Abstract
Massively multithreaded GPUs achieve high throughput by running thousands of threads in parallel. To fully utilize the hardware, workloads spawn work to the GPU in bulk by launching large tasks, where each task is a kernel that contains thousands of threads that occupy the entire GPU. GPUs face severe underutilization and their performance benefits vanish if the tasks are narrow, i.e., they contain This paper presents Pagoda, a runtime system that virtualizes GPU resources, using an OS-like daemon kernel called MasterKernel. Tasks are spawned from the CPU onto Pagoda as they become available, and are scheduled by the MasterKernel at the warp granularity. Experimental results demonstrate that Pagoda achieves a geometric mean speedup of 5.70x over PThreads running on a 20-core CPU, 1.51x over CUDA-HyperQ, and 1.69x over GeMTC, the state-of- the-art runtime GPU task scheduling system.Keywords
This publication has 28 references indexed in Scilit:
- Kernelet: High-Throughput GPU Kernel Executions with Dynamic Slicing and SchedulingIEEE Transactions on Parallel and Distributed Systems, 2013
- Multi-tenancy on GPGPU-based servers2013
- Fast concurrent queues for x86 processorsACM SIGPLAN Notices, 2013
- A virtual memory based runtime to support multi-tenancy in clusters with GPUs2012
- StarPU: a unified platform for task scheduling on heterogeneous multicore architecturesConcurrency and Computation: Practice and Experience, 2010
- StreamIt: A Language for Streaming Applications2002
- Optimal Use of Mixed Task and Data Parallelism for Pipelined ComputationsJournal of Parallel and Distributed Computing, 2000
- Dynamic storage allocation: A survey and critical review1995
- The Multifrontal Method for Sparse Matrix Solution: Theory and PracticeSIAM Review, 1992
- A fast storage allocatorCommunications of the ACM, 1965