ATS-6 Benchmarks. ATTENTION: This page is a work in progress and nothing is considered to be final

Getting Started

• 1. Introduction
  • 1.1. Benchmark Overview
    • 1.1.1. Mod/Sim Priority 1
    • 1.1.2. Mod/Sim Priority 2
    • 1.1.3. AI benchmarks
  • 1.2. Run Rules Synopsis
  • 1.3. Approvals

Mod/Sim Priority 1

• 1. Kripke
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. Problems
    • 1.2.2. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Validation
  • 1.7. Example Scalability Results
  • 1.8. Memory Usage
  • 1.9. Strong Scaling
  • 1.10. Weak Scaling on El Capitan
  • 1.11. References
• 2. Laghos
  • 2.1. Purpose
  • 2.2. Characteristics
    • 2.2.1. Problems
    • 2.2.2. Figure of Merit
  • 2.3. Source code modifications
  • 2.4. Building
    • 2.4.1. Metis (required)
    • 2.4.2. Umpire (optional)
    • 2.4.3. RAJA (required)
    • 2.4.4. Hypre (required)
    • 2.4.5. MFEM (required)
    • 2.4.6. Laghos (required)
  • 2.5. Running
  • 2.6. Validation
  • 2.7. Example Scalability Results
  • 2.8. Memory Usage
  • 2.9. Strong Scaling on El Capitan
  • 2.10. Weak Scaling on El Capitan
  • 2.11. References
• 3. RAJA Performance Suite
  • 3.1. Purpose
  • 3.2. Characteristics
    • 3.2.1. Problems
    • 3.2.2. Figure of Merit
  • 3.3. Source code modifications
  • 3.4. Building
    • 3.4.1. Getting the code
    • 3.4.2. Configuration and compilation
    • 3.4.3. MI300A architecture
    • 3.4.4. H100 architecture
  • 3.5. Running
  • 3.6. Validation
  • 3.7. Example Benchmark Results
    • 3.7.1. AMD MI300A throughput results (Priority 1 kernels)
    • 3.7.2. NVIDIA H100 throughput results (Priority 1 kernels)
  • 3.8. Memory Usage
  • 3.9. Strong Scaling on El Capitan
  • 3.10. Weak Scaling on El Capitan
  • 3.11. References
• 4. Branson
  • 4.1. Purpose
  • 4.2. Characteristics
    • 4.2.1. Problems
    • 4.2.2. Figure of Merit
  • 4.3. Source code modifications
  • 4.4. Building
  • 4.5. Running
  • 4.6. Validation
  • 4.7. Example Scalability Results
  • 4.8. Memory Usage
  • 4.9. Strong Scaling on El Capitan
  • 4.10. Weak Scaling on El Capitan
  • 4.11. References
• 5. SPARTA
  • 5.1. Purpose
  • 5.2. Characteristics
    • 5.2.1. Application Version
    • 5.2.2. Problem
    • 5.2.3. Figure of Merit
    • 5.2.4. Correctness
  • 5.3. System Information
  • 5.4. Building
    • 5.4.1. Generic
    • 5.4.2. El Capitan
  • 5.5. Running
    • 5.5.1. El Capitan
  • 5.6. Verification of Results
    • 5.6.1. Compute Figure of Merit
    • 5.6.2. El Capitan - Single Node
    • 5.6.3. El Capitan - Many Nodes
  • 5.7. References

Mod/Sim Priority 2

• 1. AMG2023
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. Problems
    • 1.2.2. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Validation
  • 1.7. Example Scalability Results
  • 1.8. Memory Usage
  • 1.9. Strong Scaling on El Capitan
  • 1.10. V-100
  • 1.11. Multi-node scaling on Crossroads
  • 1.12. References
• 2. LAMMPS ACE
  • 2.1. Purpose
  • 2.2. Characteristics
    • 2.2.1. Application Version
    • 2.2.2. Problem
    • 2.2.3. Figure of Merit
    • 2.2.4. Correctness
  • 2.3. Source Code Modifications
  • 2.4. System Information
  • 2.5. Building
    • 2.5.1. Generic
    • 2.5.2. El Capitan
  • 2.6. Running
    • 2.6.1. El Capitan
  • 2.7. Verification of Results
    • 2.7.1. Compute Figure of Merit
    • 2.7.2. El Capitan - Single Node
    • 2.7.3. El Capitan - Many Nodes
  • 2.8. References
• 3. Remhos
  • 3.1. Purpose
  • 3.2. Characteristics
    • 3.2.1. Problems
    • 3.2.2. Figure of Merit
  • 3.3. Source code modifications
  • 3.4. Building
  • 3.5. Running
  • 3.6. Validation
  • 3.7. Example Scalability Results
  • 3.8. Memory Usage
  • 3.9. Strong Scaling on El Capitan
  • 3.10. Weak Scaling on El Capitan
  • 3.11. References
• 4. MiniEM
  • 4.1. Purpose
  • 4.2. Characteristics
    • 4.2.1. Problems
    • 4.2.2. Figure of Merit
  • 4.3. Source code modifications
  • 4.4. Building
  • 4.5. Running
  • 4.6. Validation
  • 4.7. Example Scalability Results
  • 4.8. Memory Usage
  • 4.9. Strong Scaling on El Capitan
  • 4.10. Weak Scaling on El Capitan
  • 4.11. References

AI benchmarks

• 1. ScaFFold
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. Problems
    • 1.2.2. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Run Parameters
  • 1.7. Validation
  • 1.8. Example Scalability Results
  • 1.9. Memory Usage
  • 1.10. Strong Scaling on El Capitan
  • 1.11. Weak Scaling on El Capitan
  • 1.12. References
• 2. MLPerf Llama 3.1 405B training
  • 2.1. Purpose
  • 2.2. Characteristics
    • 2.2.1. Problems
    • 2.2.2. Figure of Merit
  • 2.3. Source code modifications
  • 2.4. Building
  • 2.5. Running
  • 2.6. Validation
  • 2.7. Example Scalability Results
  • 2.8. Memory Usage
  • 2.9. Strong Scaling on El Capitan
  • 2.10. Weak Scaling on El Capitan
  • 2.11. References

Storage benchmarks

• 1. IO500
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. IOR
    • 1.2.2. MDTEST
    • 1.2.3. FIND
    • 1.2.4. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Validation
  • 1.7. Example Scalability Results
  • 1.8. References
• 2. MLPerf Storage
  • 2.1. Purpose
  • 2.2. Characteristics
    • 2.2.1. Problems
    • 2.2.2. Figure of Merit
  • 2.3. Source code modifications
  • 2.4. Building
  • 2.5. Running
  • 2.6. Example Scalability Results
  • 2.7. Memory Usage
  • 2.8. Strong Scaling on El Capitan
  • 2.9. Weak Scaling on El Capitan
  • 2.10. References

Network benchmarks

• 1. Phloem
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. Problems
    • 1.2.2. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Validation
  • 1.7. Example Scalability Results
  • 1.8. Memory Usage
  • 1.9. Strong Scaling on El Capitan
  • 1.10. Weak Scaling on El Capitan
  • 1.11. References
• 2. OSU Microbenchmarks
  • 2.1. Purpose
  • 2.2. Characteristics
    • 2.2.1. Problem
  • 2.3. Source code modifications
  • 2.4. Building
  • 2.5. Running
  • 2.6. Scaling on El Capitan
  • 2.7. References
• 3. Sandia Microbenchmarks - Message rate
  • 3.1. Purpose
  • 3.2. Characteristics
    • 3.2.1. Problem
    • 3.2.2. Figure of Merit
  • 3.3. Source code modifications
  • 3.4. Building
  • 3.5. Running
  • 3.6. Validation
  • 3.7. Example Scalability Results
  • 3.8. Memory Usage
  • 3.9. Strong Scaling on El Capitan
  • 3.10. Weak Scaling on El Capitan
  • 3.11. References
• 1. IO500
  • 1.1. Purpose
  • 1.2. Characteristics
    • 1.2.1. IOR
    • 1.2.2. MDTEST
    • 1.2.3. FIND
    • 1.2.4. Figure of Merit
  • 1.3. Source code modifications
  • 1.4. Building
  • 1.5. Running
  • 1.6. Validation
  • 1.7. Example Scalability Results
  • 1.8. References

Appendices

• 1. Project Documentation
  • 1.1. Contributing
  • 1.2. Dependencies
  • 1.3. Building
  • 1.4. Plot Generation and Table Inclusion
    • 1.4.1. Creating a Table from CSV
    • 1.4.2. Creating a Figure with Gnuplot
    • 1.4.3. Creating a Figure with Matplotlib
    • 1.4.4. Ignore Relevant Files
  • 1.5. Build Script
• 2. Reference Systems
  • 2.1. El Capitan
    • 2.1.1. Node Architecture
  • 2.2. CTS-2
    • 2.2.1. Node Architecture
  • 2.3. H100 System
    • 2.3.1. Node Architecture