Title | ||
---|---|---|
POSTER: Accelerate GPU Concurrent Kernel Execution by Mitigating Memory Pipeline Stalls |
Abstract | ||
---|---|---|
In this study, we demonstrate that the performance may be undermined in the state-of-the-art intra-SM sharing schemes for concurrent kernel execution (CKE) on GPUs, due to the interference among concurrent kernels. We highlight that cache partitioning techniques proposed for CPUs are not effective for GPUs. Then we propose to balance memory accesses and limit the number of inflight memory instructions issued from concurrent kernels to reduce memory pipeline stalls. Our proposed schemes significantly improve the performance of two state-of-the-art intra-SM sharing schemes, Warped-Slicer and SMK. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1109/PACT.2017.30 | 2017 26th International Conference on Parallel Architectures and Compilation Techniques (PACT) |
Keywords | Field | DocType |
GPUs,concurrent kernel execution,memory pipeline,memory subsystem | Kernel (linear algebra),Resource management,Pipeline transport,Computer science,Cache,Parallel computing,Real-time computing,Interference (wave propagation),Throughput,Benchmark (computing) | Conference |
ISSN | ISBN | Citations |
1089-795X | 978-1-5090-6765-7 | 2 |
PageRank | References | Authors |
0.36 | 4 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hongwen Dai | 1 | 28 | 3.14 |
Zhen Lin | 2 | 35 | 4.21 |
Chao Li | 3 | 132 | 6.04 |
chen zhao | 4 | 15 | 10.09 |
Fei Wang | 5 | 203 | 40.33 |
Nanning Zheng | 6 | 3975 | 329.18 |
Huiyang Zhou | 7 | 994 | 63.26 |