demo/README.md

This is a demo of the toolchain. The qalgo folder has the quantum algorithm to be benchmarked with folders:

• gen_files: This folder has the main machinery that generates the associated run files given the qasm files, job scripts and scope files that scope into environment of the container
• qasm_files: The quantum algorithm represented using the QASM instruction set
• qasm_parser: The parser that translates the QASM instruction into the instruction of the simulation package
• simulation package folder: The folder of the simulation package with the associated run files and job scripts

To generate the associated files, in this instance the qiskit files associated with the Heisenberg dynamics:

1. Create qiskit folder in qalgo folder

• mkdir -v qiskit

2. Change into the gen_files folder and run the file

• python3 create_dir_struct.py (allows to set the options inside the file)

3. In the qiskit folder the following folders are generated:

• bash_scripts_st_dp: bash scripts that scope the run files into the environment of the container
• data_st_dp: data folder to store the time statistics
• error_st_dp: error folder to store the errors of the job
• output_st_dp: output folder to store the output of the job
• job_scripts_st_dp: job scripts folder that has the jobs to be executed on the cluster (tailored for PSI cluster)
• run_files_st_dp: translated run files from qasm to qiskit instruction set

4. run_jobs_st_dp.sh: pushes all the jobs in the bash_scripts_st_dp onto the cluster

=============================================================================================================================

Note:

1. The above instructions are tailored for a demo i.e., a minimal reproduction of the toolchain and its files (for the entire version: please see the files in the folder toolchain)

2. 1. The singularity containers are not contained in this repo but are on the zenodo repo. These can be accessed using the following commands or as in the job scripts:
   • singularity shell qucos_sim.sif

3. The containerized image has conda environments for all the packages (also in bash_script_sample.sh in zenodo repo) that is as follows:

   #!/bin/bash -l
   . /app/etc/profile.d/conda.sh
   conda activate sim_package
   python3 /home/<translated_run_file.py>

   # Available conda environments on CPU singularity image: qucos_sim.sif
   # Can be activated by issuing the commands as above, replace sim_package
   # by the package of choice as below:
   # sim_braket
   # sim_cirq
   # sim_hybridq
   # sim_myqlm
   # sim_myqlm_cpp
   # sim_pennylane
   # sim_pennylane_l
   # sim_projectq
   # sim_qibo
   # sim_qibojit
   # sim_qiskit
   # sim_qpanda
   # sim_qsimcirq
   # sim_qulacs
   # sim_svsim

   # Available conda environments on GPU singularity image: qucos_gpu_sim.sif
   # Can be activated by issuing the commands as above, replace sim_package
   # by the package of choice as below:
   # sim_hybridq_gpu
   # sim_pennylane_l_gpu
   # sim_qcgpu_gpu
   # sim_qibojit_gpu
   # sim_qiskit_gpu
   # sim_qpanda_gpu
   # sim_qsimcirq_gpu
   # sim_qulacs_gpu
   # sim_svsim_gpu

4. The use of containerized image implies that the packages installed in the container do not need additional dependencies and can be used in a straight out-of-box fashion. However to check the dependencies one can issue the command pip list once the corresponding conda environment is activated (conda environments listed above).