Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization - Archive ouverte HAL
Conference Papers Year : 2022

Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization

Mihail Popov
  • Function : Author
  • PersonId : 1129906
Akash Dutta
  • Function : Author
  • PersonId : 1129907
Emmanuelle Saillard
Ali Jannesari
  • Function : Author
  • PersonId : 1129908

Abstract

There is a large space of NUMA and hardware prefetcher configurations that can significantly impact the performance of an application. Previous studies have demonstrated how a model can automatically select configurations based on the dynamic properties of the code to achieve speedups. This paper demonstrates how the static Intermediate Representation (IR) of the code can guide NUMA/prefetcher optimizations without the prohibitive cost of performance profiling. We propose a method to create a comprehensive dataset that includes a diverse set of intermediate representations along with optimum configurations. We then apply a graph neural network model in order to validate this dataset. We show that our static intermediate representation based model achieves 80% of the performance gains provided by expensive dynamic performance profiling based strategies. We further develop a hybrid model that uses both static and dynamic information. Our hybrid model achieves the same gains as the dynamic models but at a reduced cost by only profiling 30% of the programs.
Fichier principal
Vignette du fichier
Learning_Intermediate_Representations_using_Graph_Neural_Networks_for_NUMA_and_Prefetchers_Optimization.pdf (986.03 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03603118 , version 1 (09-03-2022)

Identifiers

  • HAL Id : hal-03603118 , version 1

Cite

Ali Tehranijamsaz, Mihail Popov, Akash Dutta, Emmanuelle Saillard, Ali Jannesari. Learning Intermediate Representations using Graph Neural Networks for NUMA and Prefetchers Optimization. IPDPS 2022 - 36th IEEE International Parallel & Distributed Processing Symposium, May 2022, Lyon / Virtual, France. ⟨hal-03603118⟩
115 View
135 Download

Share

More