‘Quantum Imaginary Time Evolution’, an Evolved Classical Computing Algorithm for Quantum Computing

A California Institute of Technology team of researchers demonstrated calculation of Hamiltonians on quantum computers using an evolved classical algorithm.  Adapting the classical computing algorithm known as ‘imaginary time evolution’, the team, led by Garnet Chan of CALTECH, applied the concept to quantum systems.  Aptly named, the ‘quantum imaginary time evolution algorithm’, it stands to mitigate the detrimental effects of environmental noise on quantum bits, or qubits. 


Environmental noise in a quantum system leads to decoherence of the quantum state of the qubits.  When this occurs, quantum calculations and their results are lost.  At the present state of quantum computers, coherence times are not sufficient to capture the quantum information efficiently.  By determining the Hamiltonians of the quantum system, the team has been able to alter the algorithm to enable readout of the quantum information prior to the qubit’s decoherence. 

The Tools Used

Using a Python quantum computing program library, pyQuil, the CALTECH team was able to interface with Rigetti’s quantum virtual machine and the Aspen-1 quantum processor to implement the algorithm.  Use of pyQuil permitted interacting with the quantum systems in such a way as to produce noise models and readout of errors.  These readout errors may be mitigated by applying noise models to the system. 

Known Issues Beyond Noise

The team states not all sources of error are captured.  A known issue with the Rigetti quantum virtual machine is the inability to determine qubit-to-qubit noise, or cross-talk.  However, “compared with their classical counterparts, [the algorithms] require exponentially less space and time per iteration…”

Two other methods, phase estimation and ‘the variational approach’ have inherent failings.  Phase estimation is a complex computation taking too long to produce results with today’s nascent quantum computing environments.  The variational approach lacks accuracy in results generated.


Successful application of the algorithm is seen as having lasting effects on the current and near-future of quantum computing and machine learning.  The CALTECH team has aimed to produce a fast and accurate algorithm to compute and apply the Hamiltonian of a quantum computing system.  In doing so, results from computations on quantum computing systems are improved.  This is all in the attempt to reduce error, the bane of current quantum systems.

Because Quantum is Coming.  Qubit.

Read More [PDF]…