Skip to Main content Skip to Navigation
Conference papers

Hedgehog: a Performance-Oriented General Purpose Library for Multi-GPU Systems

Abstract : We present Hedgehog, a general-purpose library for taking advantage of powerful compute nodes, multicore CPUs, and multiple GPUs. The novel aspects of Hedgehog are: (1) its explicit representation of a program as a dataflow graph, (2) its pure dataflow-driven scheduling, (3) its maintenance of a computation's localized state via state managers, and (4) its fine control of memory via memory managers. This dataflow approach results in extremely low overhead for task executions (< 1 microsecond) and no-cost profiling at the task level. This allows us to prototype operations that compare favorably with leading libraries such as cuBLAS-XT.
Complete list of metadata
Contributor : Bruno Bachelet <>
Submitted on : Tuesday, October 27, 2020 - 5:45:16 PM
Last modification on : Wednesday, February 24, 2021 - 4:24:03 PM


  • HAL Id : hal-02981031, version 1



Alexandre Bardakoff, Timothy Blattner, Bruno Bachelet, Walid Keyrouz, Loïc Yon. Hedgehog: a Performance-Oriented General Purpose Library for Multi-GPU Systems. GPU Technology Conference (GTC), NVIDIA, Mar 2020, San Diego, United States. ⟨hal-02981031⟩



Record views