Installation

DiSLOUTrajectories.jl requires Julia 1.10 or later and QuantumToolbox 0.50 or 0.51. To install it, run in an interactive session (REPL):

using Pkg
Pkg.add("DiSLOUTrajectories")

Alternatively, this can also be done in Julia's Pkg REPL by pressing the key ] in the REPL to use the package mode, and then type the following command:

(1.10) pkg> add DiSLOUTrajectories

More information about Julia's package manager can be found at Pkg.jl.

To load the package and check the version information, use either DiSLOUTrajectories.versioninfo() or DiSLOUTrajectories.about(), namely

using DiSLOUTrajectories
DiSLOUTrajectories.versioninfo()
DiSLOUTrajectories.about()

QuantumToolbox.jl

DiSLOUTrajectories.jl is built upon QuantumToolbox.jl, which is a cutting-edge Julia package designed for quantum physics simulations, closely emulating the popular Python QuTiP package. It provides many useful functions to create arbitrary quantum states and operators which can be combined in all the expected ways. It uniquely combines the simplicity and power of Julia with advanced features like GPU acceleration and distributed computing, making simulation of quantum systems more accessible and efficient.

GPUs

dislou_solve keeps the array types of its inputs, so GPUs need no extension. With CUDA.jl, for example, QuantumToolbox's cu moves the operators and the initial state to the GPU:

using CUDA
sol = dislou_solve(cu(H), cu(ψ0), tlist, cu.(c_ops); gauge_set)

The eigendecomposition of each gauge and the propagation of the trajectories then run on the GPU. This relies on the dense linear algebra of the GPU package (eigen, qr and triangular solves) and on GPU support in QuantumToolbox's mcsolve. It is not tested in CI, which has no GPU.

Continue with the quick start.