Quantum Adiabatic Computing simulator

AdiaComput.jl is a package for simulating quantum adiabatic computing implemented in Julia. The design of AdiaComput.jl is aimed to implementing an efficient, pluggable simulator which allows users create their own adiabatic computing routine. Current AdiaComput.jl only supports CPU computing in the simulation. GPU branch of the simulator is old and needs to be further developed.

• QuComputStates.jl
• QuDynamics.jl
• JuMP.jl
• QuSAT.jl
• QuBase.jl
• Ipopt.jl

Use the following commands in Julia to install dependencies.

Pkg.clone("https://github.com/Roger-luo/QuComputStates.jl")
Pkg.clone("https://github.com/JuliaQuantum/QuDynamics.jl")
Pkg.add("JuMP")
Pkg.clone("https://github.com/Roger-luo/QuSAT.jl")
Pkg.clone("https://github.com/JuliaQuantum/QuBase.jl")
Pkg.add("Ipopt")

The package has not been submitted into METADATA.jl, therefore, if you want to use this package, you need to use Pkg.clone

Pkg.clone("https://github.com/Roger-luo/AdiaComput.jl")

For adiabatic computing, you need to generate your own problem Hamiltonian, denoted as HP in following codes.

For example, we could use the build-in QuSAT.jl interface to generate a 4-qubit Hamiltonian for exact cover problem.

using QuDynamics
using AdiaComput
using QuSAT
using QuBase
using QuComputStates

ins,ans = generate(4)
pH = QuArray(pHamiltonian(ins,4),(comput_basis(4),comput_basis(4)))

Then you will need to create an AQC obeject, there are few key words that you could tweak at the beginning.

• maxtime max evolution time.
• method method for Shrodinger equation solver. All the method in QuDynamics.jl is allowed.
  • ODE solvers QuODE45,QuODE23s,QuODE78
  • Expm solvers QuExpokit,QuExpmV

aqc = AQC(pH,4;maxtime=100)

use |> operator to insert routines into your simulated quantum adiabatic computer. In this example, we simply simulate the normal routine of an adiabatic computing.

(0.0:1e-2:100,QuODE45())|>aqc
for (t,psi) in aqc end

The tuple on the left of |> consist of two parts

• time list for your propagator (evolution)
• your propagator/evolution routine

The AQC's states is stored in aqc.state, you could check the success probability of your evolution.

p = norm(aqc.state[ans[1]+1])^2

To custome the routines, a subtype of QuDynamics.QuPropagatorMethod will need to be defined. And you will need to define your own propagate function, which returns an QuArray as next state.