FAQ

1. What commands are used to run Quantum ESPRESSO?

Here is a list of the commands available to run Quantum ESPRESSO, in alphabetical order:

2. Is Quantum ESPRESSO using MPI or OpenMP for parallel execution?

Our Quantum ESPRESSO build supports both MPI and OpenMP parallelization. The parallelization method depends on the binary you choose to run.

• To use MPI, run the standard executable, e.g. pw.x. Inductiva will take care of setting the MPI configurations automatically for you, based on the machine you selected before.
• To use OpenMP, run the executable with the _openmp suffix, e.g. pw_openmp.x.

This allows you to select the most suitable parallelization strategy for your simulation.

3. How does Inductiva implicitly set MPI configurations for me?

When you run a Quantum ESPRESSO simulation using the Inductiva API, the MPI settings are automatically configured based on the computational resources you choose. This means you don’t need to manually set MPI parameters for each command, Inductiva handles it for you.

For example:

# Allocate a machine on Google Cloud Platform
cloud_machine = inductiva.resources.MachineGroup(
    provider="GCP",
    machine_type="c2d-highcpu-16",
    spot=True)

For this machine, Inductiva will automatically apply the following MPI configurations:

• np → Number of MPI processes
• use_hwthread_cpus → Whether hyperthreading is enabled

These settings will be applied to all commands that support MPI, for example:

commands = [
    "pw.x -i Al_local_pseudo.in"
]

4. How can I explicitly configure my MPI settings?

When running Quantum ESPRESSO simulations, you can configure MPI settings by specifying the MPIConfig parameters in your computational resource.

Example:

# Allocate a machine on Google Cloud Platform
cloud_machine = inductiva.resources.MachineGroup(
    provider="GCP",
    machine_type="c2d-highcpu-4",
    spot=True)

# Configure MPI settings for the machine
mpi_config = inductiva.commands.MPIConfig(
    version="4.1.6",            # MPI version to use
    np=2,                       # Number of MPI processes
    use_hwthread_cpus=False)    # Whether to use hyperthreading

cloud_machine.set_mpi_config(
    mpi_config = mpi_config)

In this example, the specified MPI configuration will automatically be applied to all commands that support MPI.

For instance, a command like:

command = [
    "pw.x -i Al_local_pseudo.in"
]

will internally be executed as:

mpirun -np 2 --use-hwthread-cpus pw.x -i Al_local_pseudo.in

This ensures that the MPI settings are consistently applied across your Quantum ESPRESSO runs.

Still can't find what you're looking for? Contact Us