1QBit was born, tied to one quantum hardware platform
Once upon a time in 2012, the company 1QBit was founded by two quantum-interested entrepreneurs from Minor Capital VCC, in an office above a local pub in Vancouver. The entrepreneurs were likely observing the activity in nearby Burnaby, British Columbia, of a group of physicists working, on what D-Wave’s Eric Ladizinsky described as his ‘Manhattan Project’, to create the first functioning quantum computer. In the previous year: 2011, D-Wave published a landmark paper by Johnson et al, which first described D-Wave’s flux qubits.
Here is a brief story of 1QBit’s journey with its Good Chemistry spin-off, for how to meet challenges and grow and adapt in the fast changing, cutting edge, quantum technology field, if you are a quantum software company.
1QBit’s early purpose was to provide the algorithms and software for D-Wave’s unique quantum hardware. Their link with one hardware platform was strong. In the first year after 1QBit’s founding, in 2013, 1QBit received a funding seed round, a Machine Access Agreement with D-Wave, they filed their first provisional patent, and the company rolled forward on its current, decade-long path of solving optimization problems.
Figure 1. A bridge-shaped word cloud of keywords from the abstract summaries of the 24 1QBit granted patents
1QBit’s technical team steadily scaled upward too. Their optimization focus is one-quarter of the quantum market according to McKinsey’s December 2021 report Quantum computing: An emerging ecosystem and industry use cases (see their Exhibit 5 chart, teal-colored boxes).
- In 2014 – the 1QBit Team reached 30 members
- In 2015 – the 1QBit team scaled to 40 members
1QBit’s Evolution to Hardware-Agnosticism
How long did it take for 1Qbit to transition their software-hardware dependency from one quantum platform to be hardware-agnostic? Answer: Approximately five years to the Spring-Summer of 2017, based on:
- 1QBit’s earliest patent filing (keyword: Agnostic) and
- 1QBit’s June 2017 Services Web page
In 2017, 1QBit scaled to 50 team members, and by 2020, 122 Persons with 47 PhDs. The transition to an agnostic quantum-hardware perspective was clearly important for their business.
Since 2017, the company is fully-hardware agnostic, which can be seen from a summary of their 24 1QBit granted patents in Fig. 1 and their 1qloud service hardware in Fig. 2. Here are their four quantum, or quantum-inspired platforms they provided in 2020-2021 in their 1qloud service: Fujitsu Digital Annealer, NTT Optical Ising-Model Solver, Toshiba Simulated Bifurcation Machine, D-Wave Quantum Annealer. These four hardware services follow the Ising model concept. 1Qbit have since transitioned their cloud service to Microsoft Azure.
Figure 2. The 1QBit “1qloud” service offered in early 2020 until late 2021, providing: Fujitsu Digital Annealer, NTT Optical Ising-Model Solver, Toshiba Simulated Bifurcation Machine, D-Wave Quantum Annealer. 1QBit have since moved their cloud services to Microsoft Azure.
Optimization Algorithms with a Financial Use Case
Financial applications are a strength of IQBit. Herman et al.’s, 2022 paper is a good reference for Financial Use Cases, for example: Portfolio Optimization, to: compare Classical and near-term Quantum Computer solutions, to understand which quantum hardware is better suited, for which algorithms. As quantum algorithms are tied to the hardware, optimization algorithms provide a rich context from which to follow specific quantum business opportunities. The following list from Herman et al’s, 2022: A Survey of Quantum Computing for Finance provides a glimpse of the rich optimization algorithm landscape. Pay attention to the terms: Hamiltonian and Hybrid; they are key to 1QBit’s versatility and how 1QBit and its spin-off: Good Chemistry, are quantum optimization barometers for the rest of us.
Combinatorial Optimization — to minimize or maximize a function that depends on a large number of variables.
- Quantum Annealing: QA is the adiabatic evolution of a time- dependent transverse-magnetic field, Ising Hamiltonian.
- Variational Quantum Eigensolver: The VQE relies upon the Rayleigh-Ritz variational principle to find the ground state of a given problem Hamiltonian. It is a hybrid-quantum-classical algorithm, where the quantum procedure provides the state (“ansatz”) and the measurement, and the classical computer to process the measurement results and update the quantum processor.
Figure 3. Annotated Fig. 6 from McArdle et al, 2020, Quantum computational chemistry: describing the VQE hybrid process. The state preparation and measurement subroutines are performed on the quantum computer. The measured observable and parameters are fed into a classical optimization routine (green, lower panel), which outputs new values of the parameters. The new parameters are then fed back into the quantum circuit.
- Quantum Approximate Optimization Algorithm: The QAOA is a hybrid quantum-classical algorithm that also relies on the variational principle, like VQE, with the same split between quantum processor (ansatz, measurements) and classical (process and update), which is a truncated or version of the Quantum Annealing evolution to a finite number of time steps.
- Variational Quantum Imaginary Time Evolution: The VarQITE is a special case of VQE with a quantum natural gradient optimization to find the ground states in imaginary time.
- Optimization by Quantum Unstructured Search See Grover’s Quantum Algorithm Applied to Optimization by Baritompa et al, 2005
Variational circuits serve as a bridge between the classical and the quantum world, according to Nathan Killoran, from XanaduAI at 25:49 of his 2020 Quantum Algorithm talk: “Quantum machine learning on near-term devices“. Killoran uses the term: Quantum nodes to straddle the world between quantum and classical and provide a way to address some intractable problems now. The classical computer just sees a function. The quantum computer looks at the fine-grained circuit details. This is an exceptionally clear talk, from which I concluded that: every organization needs a ‘Nathan Killoran’ to explain quantum details to their deep-tech audience.
A good introduction and complement to the Portfolio Optimization topic is the 1QBit White paper: Portfolio Optimization Using Hybrid Quantum Algorithms, which solves the portfolio optimization problem: 1) Classically, 2) with the VQE, and, 3) with the QAOA, using Qiskit (IBM’s quantum SDK). They found QAOA tends to be more accurate than VQE, relative to classical. Write anish@goodchemistry.com for a copy or more information.
Figure 4. Hamiltonian sections of a variety of quantum technology modalities from the Table of Contents of: Quantum Computation and Quantum Information by Nielsen and Chuang
Mathematicians and Mathematical Physicists have a key role in quantum technology by providing and solving the Hamiltonians, that is, the function that represents the total energy of a physical quantum system. Each modality, or quantum technology, has its own way of determining a qubit. See a snapshot (Fig. 4) of the Table of Contents from the excellent text: Quantum Computation and Quantum Information by Nielsen and Chuang.
Mapping Hamiltonians to Hamiltonians and Watching them Disappear
Then you must map the Hamiltonian of the physics problem you wish to solve to the Hamiltonian of the underlying quantum hardware. Today, entire sessions of quantum computation are devoted to Hamiltonians, here, this past summer Jul 20, 2022, at TQC2022.
With that physical system represented, quantum agnostic companies like 1QBit and its spin-off: Good Chemistry, reveal the process of quantum cloud vendors removing, or abstracting out, those quantum processor details. That leaves you free to focus on the second Hamiltonian, the problem to solve. See how the VQE algorithm (Fig. 5) in Good Chemistry’s open source library Tangelo abstracts out the hardware, where Backend Options represent the type of quantum processor: Ion Trap (IonQ), or Superconducting (IBM).
Figure 5. VQE module from Tangelo: the Open-source Python Package for End-to-end Chemistry Workflows on Quantum Computers.
This is the first article in a two-part series. The second article will be published next week.
*****
Amara Graps, Ph.D. is an interdisciplinary physicist, planetary scientist, science communicator and educator and expert on all quantum technologies.