Draft NERSC-8 / Trinity Benchmarks
Check for updates to this page frequently. Benchmarks and instructions will be added as they become available.
These draft benchmark programs are under consideration for use as part of the joint NERSC / ACES NERSC-8/Trinity system procurement. There are two basic kinds of benchmarks:
MiniApplications: miniFE, miniGhost, AMG, UMT, GTC, MILC, SNAP, and miniDFT
MicroBenchmarks: Pynamic, STREAM, OMB, SMB, ZiaTest, IOR, Metabench, PSNAP, FSTest, mpimemu, and UPC_FT
The SSP is an aggregate measure based on selected runs of the MiniApplications.
Instructions: DRAFT benchmark run rules that also incorporate instructions for the NERSC capability improvement runs are now available (last updated May 20, 2013).
- SSP Brief description, sample calculation, and Hopper baseline SSP value.
- Change Log
- miniFE MiniFE mimics the finite element generation, assembly and solution for an unstructured grid problem.
- MiniGhost a Finite Difference mini-application that implements a difference stencil across a homogenous three dimensional domain.
- AMG AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids.
- UMT UMT is a 3-D, deterministic, multigroup, photon transport code for unstructured meshes.
- GTC 3D Gyrokinetic Toroidal Code
- SNAP SNAP is a proxy for the performance of a modern discrete ordinates neutral particle transport application.
- MILC MILC represents part of a set of codes used to study Lattice Quantum ChromoDynamics (QCD).
- MiniDFT MiniDFT is intended to capture the performance-critical portions of a density functional theory materials science computation.
- NPB UPC-FT This is the NAS Parallel Benchmark FFT program written in the UPC language.
- Pynamic Pynamic tests dynamic loading subsystem design and the ability to handle heavy use of dynamically linked libraries from a large Python-based scientific application.
- STREAM The STREAM benchmark is designed to measure the sustainable memory bandwidth and corresponding computation rate for four simple vector kernels
- OMB MPI Tests The OSU MicroBenchmarks carry out a variety of message passing performance tests using MPI.
- SMB The "Host Processor Overhead" and "Real World Message Rate" benchmarks from the Sandia MPI Micro-Benchmark Suite (SMB).
- PSNAP P-SNAP is a micro benchmark to measure OS impact on parallel jobs.
- ZiaTest This test executes a new proposed standard benchmark method for MPI startup that is intended to provide a realistic assessment of both launch and wireup requirements.
- mdtest Used to measure file system metadata operation rates
- IOR IOR is used for testing performance of parallel file systems using various interfaces and access patterns.
- mpimemu A simple tool that helps approximate MPI library memory usage as a function of scale.