Simultaneous Branch and Warp Interweaving for Sustained GPU Performance

Nicolas Brunie, Sylvain Collange, Gregory Diamos
ARENAIRE (Inria Grenoble Rhone-Alpes / LIP Laboratoire de l’Informatique du Parallelisme), INRIA – CNRS : UMR5668 – Universite Claude Bernard – Lyon I – Ecole Normale Superieure de Lyon
HAL : ensl-00649650, version 1, 2011


   title={Simultaneous Branch and Warp Interweaving for Sustained GPU Performance},

   author={Brunie, N. and Collange, S. and Diamos, G. and others},



Download Download (PDF)   View View   Source Source   



Single-Instruction Multiple-Thread (SIMT) micro-architectures implemented in Graphics Processing Units (GPUs) run fine-grained threads in lockstep by grouping them into so-called warps to amortize the cost of instruction fetch, decode and control logic over multiple execution units. As individual threads take divergent execution paths, their processing takes place sequentially, defeating part of the efficiency advantage of SIMD execution. We present two complementary techniques that mitigate the impact of thread divergence on SIMT micro-architectures. Both techniques relax the SIMD execution model by allowing two distinct instructions to be scheduled to disjoint subsets of the the same row of execution units, instead of one single instruction. They increase flexibility by providing more thread grouping opportunities than SIMD, while preserving the affinity between threads to avoid introducing extra memory divergence. We consider (1) co-issuing instructions from different divergent paths of the same warp and (2) co-issuing instructions from different warps. To support (1), we introduce a novel thread reconvergence technique that ensures threads are run back in lockstep at control-flow reconvergence points without hindering their ability to run branches in parallel. We propose a lane shuffling technique to allow technique (2) to benefit from inter-warp correlations in divergence patterns. The combination of all these techniques improves performance by 23% on a set of regular GPGPU applications and by 40% on irregular applications, while keeping the same instruction-fetch and processing-unit resources as the contemporary Fermi GPU architecture.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2017 hgpu.org

All rights belong to the respective authors

Contact us: