diff --git a/Tools/machines/perlmutter-nersc/perlmutter_cpu.sbatch b/Tools/machines/perlmutter-nersc/perlmutter_cpu.sbatch
index a1687b2a3b4..d13c7e3b4e5 100644
--- a/Tools/machines/perlmutter-nersc/perlmutter_cpu.sbatch
+++ b/Tools/machines/perlmutter-nersc/perlmutter_cpu.sbatch
@@ -33,6 +33,7 @@ INPUTS=inputs_small
 export SRUN_CPUS_PER_TASK=16  # 8 cores per chiplet, 2x SMP
 export OMP_PLACES=threads
 export OMP_PROC_BIND=spread
+export OMP_NUM_THREADS=${SRUN_CPUS_PER_TASK}
 
 srun --cpu-bind=cores \
   ${EXE} ${INPUTS} \
diff --git a/Tools/machines/perlmutter-nersc/perlmutter_gpu.sbatch b/Tools/machines/perlmutter-nersc/perlmutter_gpu.sbatch
index 873fd30355f..f2ea5fa3e7f 100644
--- a/Tools/machines/perlmutter-nersc/perlmutter_gpu.sbatch
+++ b/Tools/machines/perlmutter-nersc/perlmutter_gpu.sbatch
@@ -34,6 +34,7 @@ export MPICH_OFI_NIC_POLICY=GPU
 # threads for OpenMP and threaded compressors per MPI rank
 #   note: 16 avoids hyperthreading (32 virtual cores, 16 physical)
 export SRUN_CPUS_PER_TASK=16
+export OMP_NUM_THREADS=${SRUN_CPUS_PER_TASK}
 
 # GPU-aware MPI optimizations
 GPU_AWARE_MPI="amrex.use_gpu_aware_mpi=1"