Model-driven autotuning of sparse matrix-vector multiply on GPUs

Jee W. Choi, Amik Singh, Richard W. Vuduc
Georgia Institute of Technology, School of Electrical and Computer Engineering, Atlanta, Georgia, USA
Proceedings of the 15th ACM SIGPLAN symposium on Principles and practice of parallel programming, PPoPP ’10


   title={Model-driven autotuning of sparse matrix-vector multiply on GPUs},

   author={Choi, J.W. and Singh, A. and Vuduc, R.W.},

   booktitle={Proceedings of the 15th ACM SIGPLAN symposium on Principles and practice of parallel computing},





Download Download (PDF)   View View   Source Source   



We present a performance model-driven framework for automated performance tuning (autotuning) of sparse matrix-vector multiply (SpMV) on systems accelerated by graphics processing units (GPU). Our study consists of two parts. First, we describe several carefully hand-tuned SpMV implementations for GPUs, identifying key GPU-specific performance limitations, enhancements, and tuning opportunities. These implementations, which include variants on classical blocked compressed sparse row (BCSR) and blocked ELLPACK (BELLPACK) storage formats, match or exceed state-of-the-art implementations. For instance, our best BELLPACK implementation achieves up to 29.0 Gflop/s in single-precision and 15.7 Gflop/s in double-precision on the NVIDIA T10P multiprocessor (C1060), enhancing prior state-of-the-art unblocked implementations (Bell and Garland, 2009) by up to 1.8? and 1.5? for single-and double-precision respectively. However, achieving this level of performance requires input matrix-dependent parameter tuning. Thus, in the second part of this study, we develop a performance model that can guide tuning. Like prior autotuning models for CPUs (e.g., Im, Yelick, and Vuduc, 2004), this model requires offline measurements and run-time estimation, but more directly models the structure of multithreaded vector processors like GPUs. We show that our model can identify the implementations that achieve within 15% of those found through exhaustive search.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2023 hgpu.org

All rights belong to the respective authors

Contact us: