TU Darmstadt / ULB / TUbiblio

Towards compiler-aided correctness checking of adjoint MPI applications

Hück, Alexander ; Protze, Joachim ; Lehr, Jan-Patrick ; Terboven, Christian ; Bischof, Christian ; Müller, Matthias S. (2020)
Towards compiler-aided correctness checking of adjoint MPI applications.
4th International Workshop on Software Correctness for HPC Applications (Correctness). Virtual Conference (09.-19.11.2020)
doi: 10.1109/Correctness51934.2020.00010
Konferenzveröffentlichung, Bibliographie

Kurzbeschreibung (Abstract)

Algorithmic Differentiation (AD) is a set of techniques to calculate derivatives of a computer program. In C++, AD typically requires (i) a type change of the built-in double, and (ii) a replacement of all MPI calls with AD-specific implementations. This poses challenges on MPI correctness tools, such as MUST, a dynamic checker, and TypeART, its memory sanitizer extension. In particular, AD impacts (i) memory layouts of the whole code, (ii) requires more memory allocations tracking by TypeART, and (iii) approximately doubles the MPI type checks of MUST due to an AD-specific communication reversal. To address these challenges, we propose a new callback interface for MUST to reduce the number of intercepted MPI calls, and, also, improve the filtering capabilities of TypeART to reduce tracking of temporary allocations for the derivative computation. We evaluate our approach on an AD-enhanced version of CORAL LULESH. In particular, we reduce stack variable tracking from 32 million to 13 thousand. MUST with TypeART and the callback interface reduces the runtime overhead to that of vanilla MUST.

Typ des Eintrags: Konferenzveröffentlichung
Erschienen: 2020
Autor(en): Hück, Alexander ; Protze, Joachim ; Lehr, Jan-Patrick ; Terboven, Christian ; Bischof, Christian ; Müller, Matthias S.
Art des Eintrags: Bibliographie
Titel: Towards compiler-aided correctness checking of adjoint MPI applications
Sprache: Englisch
Publikationsjahr: 2020
Verlag: IEEE
Buchtitel: Proceedings of Correctness 2020: Fourth International Workshop on Software Correctness for HPC Applications
Reihe: Software Correctness for HPC Applications
Veranstaltungstitel: 4th International Workshop on Software Correctness for HPC Applications (Correctness)
Veranstaltungsort: Virtual Conference
Veranstaltungsdatum: 09.-19.11.2020
DOI: 10.1109/Correctness51934.2020.00010
Kurzbeschreibung (Abstract):

Algorithmic Differentiation (AD) is a set of techniques to calculate derivatives of a computer program. In C++, AD typically requires (i) a type change of the built-in double, and (ii) a replacement of all MPI calls with AD-specific implementations. This poses challenges on MPI correctness tools, such as MUST, a dynamic checker, and TypeART, its memory sanitizer extension. In particular, AD impacts (i) memory layouts of the whole code, (ii) requires more memory allocations tracking by TypeART, and (iii) approximately doubles the MPI type checks of MUST due to an AD-specific communication reversal. To address these challenges, we propose a new callback interface for MUST to reduce the number of intercepted MPI calls, and, also, improve the filtering capabilities of TypeART to reduce tracking of temporary allocations for the derivative computation. We evaluate our approach on an AD-enhanced version of CORAL LULESH. In particular, we reduce stack variable tracking from 32 million to 13 thousand. MUST with TypeART and the callback interface reduces the runtime overhead to that of vanilla MUST.

Freie Schlagworte: adjoint MPI, correctness, type mismatch, algorithmic differentiation
Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Scientific Computing
Zentrale Einrichtungen
Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ)
Zentrale Einrichtungen > Hochschulrechenzentrum (HRZ) > Hochleistungsrechner
Hinterlegungsdatum: 18 Jan 2021 09:30
Letzte Änderung: 15 Jun 2021 13:54
PPN:
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen