An IPC-based Dynamic Cooperative Thread Array Scheduling Scheme for GPUs (GPU 성능향상을 위한 IPC 기반 동적 CTA 스케줄링 기법 연구)

손동오; Jong Myon Kim (김종면); Cheol Hong Kim (김철홍)

An IPC-based Dynamic Cooperative Thread Array Scheduling Scheme for GPUs

Journal of The Korea Society of Computer and Information
Abbr : JKSCI
2016, 21(2), pp.9-16
Publisher : The Korean Society Of Computer And Information
Research Area : Engineering > Computer Science

손동오 ¹, Jong Myon Kim ², Cheol Hong Kim ¹

¹전남대학교
²울산대학교

Accredited

ABSTRACT

Recently, many research groups have focused on GPGPUs in order to improve the performance of computing systems. GPGPUs can execute general-purpose applications as well as graphics applications by using parallel GPU hardware resources. GPGPUs can process thousands of threads based on warp scheduling and CTA scheduling. In this paper, we utilize the traditional CTA scheduler to assign a various number of CTAs to SMs. According to our simulation results, increasing the number of CTAs assigned to the SM statically does not improve the performance. To solve the problem in traditional CTA scheduling schemes, we propose a new IPC-based dynamic CTA scheduling scheme. Compared to traditional CTA scheduling schemes, the proposed dynamic CTA scheduling scheme can increase the GPU performance by up to 13.1%.

KEYWORDS

General Purpose computation on the Graphics Processing Unit, Cooperative Thread Array Scheduling Schemes, Performance, Parallelism

Citation status

* References for papers published after 2023 are currently being built.

[confproc] V. Agarwal / 2000 / Clock rate versus IPC : the end of the road for conventional microArchitectures / Proceedings of 27th International Symposium on Computer Architecture : 248~259

[confproc] Buck / 2004 / Brook for GPUs: stream computing on graphics hardware / Proceedings of 31th Annual Conference on Computer Graphics : 777~786

[web] / NVIDIA CUDA Programming / http://www.nvidia.com/object/cuda_home_new.html

[web] / OpenCL / http://www.khronos.org/opencl/

[web] / ATI Streaming / http://www.amd.com/stream

[web] / General-purpose computation on graphics hardware / http://www.gpgpu.org

[confproc] I. A. Buck / 2007 / Programming CUDA / Supercomputing 2007 Tutorial Notes

[confproc] A. Jog / 2013 / OWL:Cooperative Thread Array Aware Scheduling Techniques for Improving GPGPU Performance / Proceedings of the International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) : 395~406

[confproc] V. Narasiman / 2011 / Improving GPU Performance via Large Warps and Two-Level Warp Scheduling / Proceedings of the International Symposium on Microarchitecture (MICRO) : 308~317

[confproc] T. G. Rogers / 2012 / Cache-Conscious Wavefront Scheduling / Proceedings of the International Symposium on Microarchitecture (MICRO) : 78~85

[confproc] V. W. Lee / 2010 / Debunking the 100X GPU vs. CPU Myth: An Evaluation of Throughput Computing on CPU and GPU / proceedings of International Symposium on Computer Architecture : 451~460

[web] / NVIDIA’s Next Generation CUDA Compute Architecture:Fermi / www.nvidia.com/content/pdf/fermi_white_papers/nvidia_fermi_compute_architecture_whitepaper.pdf

[confproc] J. E. Thornton / 1964 / Parallel operation in the control data 6600 / AFIPS Proceedings of FJCC, Part.2 26 : 33~40

[confproc] I. Buck / 2004 / Brook for GPUs: stream computing on graphics hardware / Proceedings of Conference on Computer Graphics and Interactive Techniques (SIGGRAPH) : 777~786

[confproc] M. Lee / 2014 / Improving GPGPU Resource Utilization Through Alternative Thread Block Scheduling / Proceedings of the International Symposium on High Performance Computer Architecture (HPCA) : 260~271

[confproc] A. Bakhoda / 2009 / Analyzing CUDA Workloads Using a Detailed GPU Simulator / Proceedings of 9th International Symposium on Performance Analysis of Systems and Software : 163~174

[confproc] S. Li / 2009 / McPAT: an integrated power, area, and timing modeling framework for multicore and manycore architectures / Proceedings of the International Symposium on Microarchitecture : 469~480

[confproc] J. Leng / 2013 / GPUWattch : Enabling Energy Optimizations in GPGPUs / Proceedings of the International Symposium Computer Architecture : 487~498

[web] / CUDA SDK / http://developerdownload.nvidia.com/compute/cuda/sdk/website/samples.html

[confproc] S. Che / 2009 / Rodinia: A Benchmark Suite for Heterogeneous Computing / Proceedings of the International Symposium on Workload Characterization (IISWC) : 44~54

This paper was written with support from the National Research Foundation of Korea.

KJCKorea
Journal Central

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65