Frontier (OLCF)

The Frontier cluster (see: Crusher) is located at OLCF. Each node contains 4 AMD MI250X GPUs, each with 2 Graphics Compute Dies (GCDs) for a total of 8 GCDs per node. You can think of the 8 GCDs as 8 separate GPUs, each having 64 GB of high-bandwidth memory (HBM2E).

Introduction

If you are new to this system, please see the following resources:

Crusher user guide
Batch system: Slurm
Production directories:
- $PROJWORK/$proj/: shared with all members of a project, purged every 90 days (recommended)
- $MEMBERWORK/$proj/: single user, purged every 90 days (usually smaller quota, 50TB default quota)
- $WORLDWORK/$proj/: shared with all users, purged every 90 days (50TB default quota)
- Note that the $HOME directory is mounted as read-only on compute nodes. That means you cannot run in your $HOME. It’s default quota is 50GB.

Note: the Orion lustre filesystem on Frontier and the older Alpine GPFS filesystem on Summit are not mounted on each others machines. Use Globus to transfer data between them if needed.

Installation

Use the following commands to download the WarpX source code and switch to the correct branch:

git clone https://github.com/ECP-WarpX/WarpX.git $HOME/src/warpx

We use the following modules and environments on the system ($HOME/frontier_warpx.profile).

Listing 18 You can copy this file from Tools/machines/frontier-olcf/frontier_warpx.profile.example.

# please set your project account
#export proj=APH114-frontier

# required dependencies
module load cmake/3.23.2
module load craype-accel-amd-gfx90a
module load rocm/5.2.0  # waiting for 5.5 for next bump
module load cray-mpich
module load cce/15.0.0  # must be loaded after rocm

# optional: faster builds
module load ccache
module load ninja

# optional: just an additional text editor
module load nano

# optional: for PSATD in RZ geometry support (not yet available)
#module load cray-libsci_acc/22.06.1.2
#module load blaspp
#module load lapackpp

# optional: for QED lookup table generation support
module load boost/1.79.0-cxx17

# optional: for openPMD support
module load adios2/2.8.3
module load cray-hdf5-parallel/1.12.2.3

# optional: for Python bindings or libEnsemble
module load cray-python/3.9.13.1

# fix system defaults: do not escape $ with a \ on tab completion
shopt -s direxpand

# make output group-readable by default
umask 0027

# an alias to request an interactive batch node for one hour
#   for paralle execution, start on the batch node: srun <command>
alias getNode="salloc -A $proj -J warpx -t 01:00:00 -p batch -N 1 --ntasks-per-node=8 --gpus-per-task=1 --gpu-bind=closest"
# an alias to run a command on a batch node for up to 30min
#   usage: runNode <command>
alias runNode="srun -A $proj -J warpx -t 00:30:00 -p batch -N 1 --ntasks-per-node=8 --gpus-per-task=1 --gpu-bind=closest"

# GPU-aware MPI
export MPICH_GPU_SUPPORT_ENABLED=1

# optimize ROCm/HIP compilation for MI250X
export AMREX_AMD_ARCH=gfx90a

# compiler environment hints
export CC=$(which cc)
export CXX=$(which CC)
export FC=$(which ftn)
export CFLAGS="-I${ROCM_PATH}/include"
export CXXFLAGS="-I${ROCM_PATH}/include -Wno-pass-failed"
export LDFLAGS="-L${ROCM_PATH}/lib -lamdhip64"

We recommend to store the above lines in a file, such as $HOME/frontier_warpx.profile, and load it into your shell after a login:

source $HOME/frontier_warpx.profile

Then, cd into the directory $HOME/src/warpx and use the following commands to compile:

cd $HOME/src/warpx
rm -rf build

cmake -S . -B build -DWarpX_COMPUTE=HIP
cmake --build build -j 32

The general cmake compile-time options apply as usual.

That’s it! A 3D WarpX executable is now in build/bin/ and can be run with a 3D example inputs file. Most people execute the binary directly or copy it out to a location in $PROJWORK/$proj/.

Running

MI250X GPUs (2x64 GB)

After requesting an interactive node with the getNode alias above, run a simulation like this, here using 8 MPI ranks and a single node:

runNode ./warpx inputs

Or in non-interactive runs:

Listing 19 You can copy this file from Tools/machines/frontier-olcf/submit.sh.

#!/usr/bin/env bash

#SBATCH -A <project id>
#SBATCH -J warpx
#SBATCH -o %x-%j.out
#SBATCH -t 00:10:00
#SBATCH -p batch
# Currently not configured on Frontier:
#S BATCH --ntasks-per-node=8
#S BATCH --cpus-per-task=8
#S BATCH --gpus-per-task=1
#S BATCH --gpu-bind=closest
#SBATCH -N 20

# load cray libs and ROCm libs
#export LD_LIBRARY_PATH=${CRAY_LD_LIBRARY_PATH}:${LD_LIBRARY_PATH}

# From the documentation:
# Each Frontier compute node consists of [1x] 64-core AMD EPYC 7A53
# "Optimized 3rd Gen EPYC" CPU (with 2 hardware threads per physical core) with
# access to 512 GB of DDR4 memory.
# Each node also contains [4x] AMD MI250X, each with 2 Graphics Compute Dies
# (GCDs) for a total of 8 GCDs per node. The programmer can think of the 8 GCDs
# as 8 separate GPUs, each having 64 GB of high-bandwidth memory (HBM2E).

# note (5-16-22 and 7-12-22)
# this environment setting is currently needed on Frontier to work-around a
# known issue with Libfabric (both in the May and June PE)
#export FI_MR_CACHE_MAX_COUNT=0  # libfabric disable caching
# or, less invasive:
export FI_MR_CACHE_MONITOR=memhooks  # alternative cache monitor

# note (9-2-22, OLCFDEV-1079)
# this environment setting is needed to avoid that rocFFT writes a cache in
# the home directory, which does not scale.
export ROCFFT_RTC_CACHE_PATH=/dev/null

export OMP_NUM_THREADS=1
export WARPX_NMPI_PER_NODE=8
export TOTAL_NMPI=$(( ${SLURM_JOB_NUM_NODES} * ${WARPX_NMPI_PER_NODE} ))
srun -N${SLURM_JOB_NUM_NODES} -n${TOTAL_NMPI} --ntasks-per-node=${WARPX_NMPI_PER_NODE} \
    ./warpx inputs > output.txt

Post-Processing

For post-processing, most users use Python via OLCFs’s Jupyter service (Docs).

Please follow the same guidance as for OLCF Summit post-processing.

Known System Issues

Warning

May 16th, 2022 (OLCFHELP-6888): There is a caching bug in Libfabric that causes WarpX simulations to occasionally hang on Frontier on more than 1 node.

As a work-around, please export the following environment variable in your job scripts until the issue is fixed:

#export FI_MR_CACHE_MAX_COUNT=0  # libfabric disable caching
# or, less invasive:
export FI_MR_CACHE_MONITOR=memhooks  # alternative cache monitor

Warning

Sep 2nd, 2022 (OLCFDEV-1079): rocFFT in ROCm 5.1+ tries to write to a cache in the home area by default. This does not scale, disable it via:

export ROCFFT_RTC_CACHE_PATH=/dev/null

Warning

January, 2023 (OLCFDEV-1284, AMD Ticket: ORNLA-130): We discovered a regression in AMD ROCm, leading to 2x slower current deposition (and other slowdowns) in ROCm 5.3 and 5.4. Reported to AMD and fixed for the next release of ROCm.

Stay with the ROCm 5.2 module to avoid.