Quantum computing is emerging as the next R&D arena for the tech players. While Google is betting big on the error-correction capacities of its latest quantum computing chip, Willow, IBM is working on delivering its own “fault-tolerant” quantum computer by 2029 through collaborations with partners like Qedma, an Israeli start-up in which it also invested recently.
The start-up is known for specialising in error-mitigation software. Its main piece of software, QESEM, or Quantum Error Suppression and Error Mitigation, analyses noise patterns to suppress some classes of errors while the algorithm is running and mitigate others in post-processing.
Qedma’s Vision Of Quantum Computation
For quantum computing, Qedma has set a mission for itself: to deliver a robust, cutting-edge, and cross-platform middleware which would provide the best error reduction software solution worldwide, today and for the foreseeable future.
While the current estimates are that the long-promised quantum commercial breakthroughs are more than a decade away, Qedma strongly believes that with its error reduction solution, the promise of quantum computations will soon be unlocked and the quantum revolution will commence far earlier than expected.
Qedma’s roadmap, in alignment with the roadmaps of the major hardware manufacturers, is designed to enable the start-up’s customers to achieve initial quantum advantages for research purposes already within the next year or two, enabling them to arrive at simulations of physical systems of 50–60 qubits, which are already hard to simulate classically.
Qedma now intends to take its customers to the next level, allowing them to be the first to achieve quantum computational advantages in a variety of algorithms working on 80–100 qubits, providing them with strong advantages in quantum algorithm design and quantum simulations.
By further incorporating Qedma’s novel solution, which allows the start-up’s error reduction software to work in conjunction with error correction, Qedma will next allow its customers to be the first to derive utility out of error correction capabilities as soon as those become available on quantum hardware. This will enable these businesses to be the first to arrive at ROI quantum advantages, working on a few hundred qubits, within the next few years.
“Our software’s benefits are designed to continue to provide huge advantages further down the road and for many years to come. With our first-mover advantage, we aim for Qedma’s technology to be ready to unlock new quantum computational capabilities for our customers as soon as developments in quantum hardware occur. We believe that by sustaining our advantage in the error reduction domain, we are set to be a key enabler of the quantum computing revolution. We believe we are making great strides towards this goal,” the start-up noted, while stating, “In the long run, our ambition is to become the standard operating system for quantum computing. Our operating system will be based on the core IP being developed at Qedma today, extended to give users a variety of cutting-edge tools beyond error reduction, including a development environment, compilation, and optimised quantum code libraries. This will enable users to easily develop their Q algorithms and optimise their performance in the quantum devices they have access to.”
Here Are The Solutions
The main problem of today’s quantum processing units (QPUs) is noise, consisting, for example, of decoherence and calibration errors. These effects significantly deteriorate the accuracy of the output of already very small quantum algorithms run on current devices. To address this issue and enable effective use of quantum computers, Qedma has developed QESEM: a world-class Quantum Error Suppression and Error Mitigation software for current QPUs.
QESEM enables users to obtain highly accurate results from their quantum algorithms, even though these algorithms are being run on the currently available QPUs, which are very noisy.
Given a set of quantum circuits to be run, a set of observables to be measured, a required statistical accuracy, and a QPU-time allocation on a particular QPU, QESEM first compiles the circuits onto a set of quantum logic operations available on the QPU. These include natively available operations as well as additional operations calibrated by Qedma (available on selected QPUs). QESEM then efficiently characterises errors in circuits composed of these quantum logic operations.
Based on this characterisation, QESEM recompiles the given circuits into a set of circuits to be run on the QPU and estimates the required QPU time. Once all circuits have been run and measurement outcomes have been collected, QESEM implements classical post-processing, apart from reporting an estimate and an error bar for each observable.
QESEM outputs provide unbiased estimates for expectation values of observables. Estimation errors are predominantly statistical and can be reduced in a controllable manner by allowing additional QPU time. The solution is also known for ensuring an unparalleled trade-off between circuit volume, statistical accuracy, and QPU time, elements which are unattainable by competing methods and software packages.
QESEM is also “Application-Agnostic,” meaning it applies to any quantum circuit and is observable. Among its other abilities, the tool reliably estimates expected QPU times (the amount of time a Quantum Processing Unit spends actively processing a job), both before and at the beginning of each run, to support users in managing their quantum resources. Also, QESEM provides support for a variety of state-of-the-art hardware platforms, something that has already been demonstrated on QPUs based on cross-resonance and tunable-coupler superconducting qubits, as well as trapped ions, by leading quantum hardware manufacturers. The start-up is now carrying out demonstrations on other hardware platforms as well.
QESEM is intended for developers and users (research groups in universities and research centres working on quantum algorithms, quantum many-body physics, or quantum chemistry) of quantum applications, software, or hardware, who are interested in running quantum algorithms on quantum processing units and receiving unbiased noise-free results at otherwise inaccessible circuit volumes, with cutting-edge statistical accuracies and QPU times.
The tool has also been customised to serve users like high-performance computing data centres, which have incorporated QPUs into their classical computing hardware and enterprises that are actively developing quantum algorithms aligned with their core business challenges and are interested in becoming quantum-ready and working towards applying quantum algorithms to achieve ROI.
Next, we have Qedma’s “Characterisation Software,” which ensures fast and detailed characterisation of multi-qubit systems, something that is essential for reducing errors in quantum computers.
The solution has been tailored for quantum hardware companies seeking to improve gate fidelities and accelerate research and development efforts, along with functions like automation of device characterisation and calibration.
These clients can also increase device throughput and reduce hardware times devoted to characterisation and calibration through “Characterisation Software.” The tool has also been developed keeping in mind the requirements of quantum cloud aggregators and data centres, which seek to boost the performance of their available quantum hardware.