multiprogramming 0.0.1

A package for multiprogramming algorithm

Multiprogramming

This repository contains the implementation of our multiprogramming mapping algorithms, including SRB, qubit mapping partition (GSP and QHSP), and mapping transition. HA algorithm is also included to evalutate the independent executions.

Installation

The user can install it by cloning with git:

https://github.com/peachnuts/Multiprogramming.git

How to use?

The user can find an example in src/main.py, which includes a list of different circuits.

Another example of finding the ground state of Deuteron using VQE can be found in src/VQE/deuteron_test.py.

Here are some notes:

1. The crosstalk_properties should be obtained before the multiprogramming algorithm, using Simultaneous Randomized Benchmarking (SRB). An example of performing SRB is shown in src/crosstalk/srb_hardware.py. The optimization methods to parallelize SRB experiments of multiple CNOT pairs is shown in src/crosstalk/neighbour_pair.py. The format of the crosstalk_properties should be like, for example:

crosstalk_properties = {(2,3):{(5,8) : 0.0317} # E(g_i|g_j) = 0.0317 with g_i = CX(2,3), g_j = CX(5,8)

2. The user should create a list of circuits that are supposed to be executed simultaneously.

3. The function main.circuits_schedule is the entry to start the multiprogramming algorithm. The last parameter is the partition method selected by the user. It can be partition_hardware_heuristic (QHSP) or partition_hardware (GSP).

4. The quantum register size of the benchmarks should be equal to the hardware qubit numbers.

Contact information

If you have any questions, feel free to contact [e-mail].