3. Adaptive MPI (AMPI)

• 3.1. Introduction
  • 3.1.1. Overview
  • 3.1.2. Charm++
  • 3.1.3. AMPI
    • 3.1.3.1. MPI Standards Compliance
• 3.2. Building and Running AMPI Programs
  • 3.2.1. Installing AMPI
  • 3.2.2. Building AMPI Programs
  • 3.2.3. Running AMPI Programs
    • 3.2.3.1. Running with charmrun
    • 3.2.3.2. Running with ampirun
    • 3.2.3.3. Other options
• 3.3. Using Existing MPI Codes with AMPI
  • 3.3.1. Entry Point
    • 3.3.1.1. Fortran
    • 3.3.1.2. C or C++
  • 3.3.2. Command Line Argument Parsing
    • 3.3.2.1. Fortran
    • 3.3.2.2. C or C++
  • 3.3.3. Global Variable Privatization
    • 3.3.3.1. Manual Code Editing
    • 3.3.3.2. PIEglobals: Automatic Position-Independent Executable Runtime Relocation
    • 3.3.3.3. TLSglobals: Automatic Thread-Local Storage Swapping
    • 3.3.3.4. PiPglobals: Automatic Process-in-Process Runtime Linking Privatization
    • 3.3.3.5. FSglobals: Automatic Filesystem-Based Runtime Linking Privatization
    • 3.3.3.6. Swapglobals: Automatic Global Offset Table Swapping
    • 3.3.3.7. Source-to-Source Transformation
    • 3.3.3.8. Summary
• 3.4. Extensions
  • 3.4.1. Serialization
  • 3.4.2. Load Balancing and Migration
  • 3.4.3. Checkpointing and Fault Tolerance
  • 3.4.4. Memory Efficiency
  • 3.4.5. Compute Resource Awareness
  • 3.4.6. Charm++ Interoperation
  • 3.4.7. Sequential Re-run of a Parallel Node
  • 3.4.8. User Defined Initial Mapping
  • 3.4.9. Performance Visualization
• 3.5. Example Applications
  • 3.5.1. Mantevo project v3.0
    • 3.5.1.1. MiniFE
    • 3.5.1.2. MiniMD v2.0
    • 3.5.1.3. CoMD v1.1
    • 3.5.1.4. MiniXYCE v1.0
    • 3.5.1.5. HPCCG v1.0
    • 3.5.1.6. MiniAMR v1.0
    • 3.5.1.7. Not yet AMPI-zed (reason)
  • 3.5.2. LLNL ASC Proxy Apps
    • 3.5.2.1. LULESH v2.0
    • 3.5.2.2. AMG 2013
    • 3.5.2.3. Lassen v1.0
    • 3.5.2.4. Kripke v1.1
    • 3.5.2.5. MCB v1.0.3 (2013)
    • 3.5.2.6. Not yet AMPI-zed (reason)
  • 3.5.3. Other Applications
    • 3.5.3.1. MILC 7.0
    • 3.5.3.2. SNAP v1.01 (C version)
    • 3.5.3.3. Sweep3D
    • 3.5.3.4. PENNANT v0.8
  • 3.5.4. Benchmarks
    • 3.5.4.1. Jacobi-2D (Fortran)
    • 3.5.4.2. Jacobi-3D (C)
    • 3.5.4.3. NAS Parallel Benchmarks (NPB 3.3)
    • 3.5.4.4. NAS PB Multi-Zone Version (NPB-MZ 3.3)
    • 3.5.4.5. HPCG v3.0
    • 3.5.4.6. Intel Parallel Research Kernels (PRK) v2.16
    • 3.5.4.7. OSU Microbenchmarks
  • 3.5.5. Third Party Open Source Libraries
    • 3.5.5.1. HYPRE-2.11.1
    • 3.5.5.2. MFEM-3.2
    • 3.5.5.3. XBraid-1.1
  • 3.5.6. Other AMPI codes

[PiP2018]

Atsushi Hori, Min Si, Balazs Gerofi, Masamichi Takagi, Jai Dayal, Pavan Balaji, and Yutaka Ishikawa. 2018. Process-in-process: techniques for practical address-space sharing. In Proceedings of the 27th International Symposium on High-Performance Parallel and Distributed Computing (HPDC ‘18). ACM, New York, NY, USA, 131-143. DOI: https://doi.org/10.1145/3208040.3208045