Command-line tools

The myQLM set of Python libraries also comes with a set of command-line scripts to generate, execute and optimize circuits

Compiler

myQLM comes with a compiler and a decompiler to convert Atos Quantum Assembly Language (AQASM) files into useable Quantum Circuit for our QPUs and vice-versa. Please refer to The AQASM format to learn more about the AQASM format.

aqasm2circ

This command takes one or two arguments, an input file (i.e. AQASM file) and optionally an output file. If no output file is supplied, the output will be the input file appended with .circ.

For example, assuming the file my_prog.aqasm contains the following code:

BEGIN
qubits 2
cbits 1
H q[0]
MEAS q[0] c[0]
?c[0]: X q[1]
END

The command aqasm2circ can be used to compile this program into a useable Quantum Circuit:

aqasm2circ my_prog.aqasm compiled.circ

The file compiled.circ is a binary file describing a quantum circuit. This circuit could be load using Python:

from qat.core import Circuit

# Load compiled circuit
compiled = Circuit.load("compiled.circ")

circ2aqasm

This is the main decompiler which will take a Quantum Circuit file and return the resulting Atos Quantum Assembly Language (AQASM) file used to generate the circuit. The command takes one or two arguments, an input file and optionally an output file. If no output file is supplied the output will be the input file appended with .aqasm.

For example, considering the file compiled.circ of the previous section, the following command can be used to generate an AQASM file equivalent to my_prog.aqasm

circ2aqasm compiled.circ my_prog_2.aqasm

Simulation / Execution commands

myQLM provides qat-jobrun and qat-batchrun to execute any Job or Batch using any QPU. Moreover, plugins coulb be linked to this QPU direclty from the command line. Moreover, the QLM provides tools to display results of a simulation / execution.

qat-jobrun and qat-batchrun

This command will deserialize a job (or a batch) and run it via some execution stack.

QPUs and Plugins are referred to using the namespace:class_name convention.

For instance, the following Python code generate a Batch and a Job using the circuit compiled with aqasm2circ. This batch and job will be then simulated using qat-batchrun and qat-jobrun.

from qat.core import Circuit, Batch

# Generate a job
circuit = Circuit.load("compiled.circ")
job = circuit.to_job()
job.dump("compiled.job")

# Generate a batch
batch = Batch(jobs=[job, job, job])
batch.dump("compiled.batch")

The qat-batchrun or qat-jobrun commands can be used to simulate this circuit. The following command call PyLinalg to simulate the batch and display the output of the simulation in the stdout.

qat-batchrun --qpu qat.pylinalg:PyLinalg compiled.batch
qat-jobrun --qpu qat.pylinalg:PyLinalg compiled.job

It is possible to dump the results in a binary file instead of having to parse the stdout

qat-batchrun --qpu qat.pylinalg:PyLinalg compiled.batch -o batchout.res
qat-jobrun --qpu qat.pylinalg:PyLinalg compiled.batch -o jobout.res

qat-resprint

Naturally, there is a command to display a dumped result in the stdout. This command can be used to display the result of qat-jobrun or qat-batchrun.

For instance, to display the result of the previous simulation:

qat-resprint jobout.res -N 2
qat-resprint -b batchres.res -N 2

qat-circstat

This command displays some statistics about a serialized circuit. The output is straightforward to interpretate:

qat-circstat compiled.circ