OpenQMC-simulator

QMC implementation for open quantum dynamics

Basic Usage

Basic working examples corresponding to the circuits simulated in are provided in the examples folder. Circuits and noise models must be constructed exclusively using HOQST, as this is the only supported method for defining simulation settings.

Download the Sysimage Asset

Download the asset (e.g. openqmc-simulator-macos.zip) from the GitHub release and extract it. The extracted file is the compiled simulator.

Launch Julia Using the Sysimage

With your standard Julia installation and instantiation, start a new session that uses the sysimage. For example, on macOS:

export JULIA_NUM_THREADS=8
julia --project=@. --sysimage compiled/lib/julia/sys.dylib

Once the REPL starts with the custom sysimage, you can run the GHZ state preparation simulation script (or other examples) from the REPL.

julia> include("examples/GHZ_state_preparation.jl")
  4.021899 seconds (9.36 M allocations: 483.778 MiB, 1.97% gc time, 77.10% compilation time)
QMCState with 413 occupied states

The output is a sparse representation of the density matrix, $\rho$, where only statistically significant states (elements in the density matrix) are stored. You can access a specific state ID using:

julia> ρ(ID::Int)

For example, ID = 1 state corresponds to the (1,1) element of the density matrix. To view the full list of states:

julia> ρ.pop_list

For a full-state tomography, use:

julia> reconstruct_state(ρ)
32×32 SparseMatrixCSC{ComplexF64, Int64} with 412 stored entries:
⎡⣿⣿⢕⢟⠑⢄⢕⢟⠟⠅⠁⠀⠟⠅⣵⣿⎤
⎢⣵⢕⣿⣿⠑⠄⠑⠄⠁⠀⠁⠀⣵⢕⢕⣵⎥
⎢⠑⢄⠑⠄⣿⣿⠕⠟⠑⠄⠑⠄⠑⠄⠑⠄⎥
⎢⣵⢕⠑⠄⣵⠅⢟⣵⣵⢕⠑⠄⠑⠄⢕⣵⎥
⎢⠑⠄⠀⠀⠑⠄⠅⠁⣿⣿⠑⢟⠑⢄⠅⠟⎥
⎢⠀⠀⠁⠀⠁⠀⠁⠀⣵⢅⢟⣥⠀⠀⠁⠀⎥
⎢⢟⠵⠕⠟⠑⠄⠑⠄⠑⢄⠁⠀⢟⣵⠕⢟⎥
⎣⣵⣿⢟⣵⠑⠄⢕⣵⣥⠅⠁⠀⡵⢅⣿⣵⎦

Note that full-state tomography can be computationally expensive for large systems. Therefore, it's often more efficient to perform local measurements that access specific elements in the density matrix.