The state of a qubit, the unit of quantum information, is complicated. It can be 1, it can be 0, or it can be something in between described with vectors and complex amplitudes. Then, of course, if there is more than one qubit, we need to talk about entanglement and superposition.
Let's switch to an easier discussion about the state of quantum computing.
Last week in Santa Clara, CA, QCWare hosted its annual quantum computing conference, Q2B- Practical Quantum Computing. Unlike last year's virtual conference, this was in person and had over 600 attendees. Q2B was well attended by the leading companies and academia in the quantum computing industry. Members representing all aspects of the quantum computing community were in attendance, including computing providers, software providers, and researchers. Here are my observations and takeaways from the conference.
Quantum Computing Hardware
Great strides have been made in the development of quantum computers, with many systems featuring 50 to 100+ qubits. The greater the number of qubits, the more complex the quantum problems can be solved. Although the goal is a fault-tolerant 1k+ qubit and beyond quantum computers, where qubits are immune to noise and can sustain their quantum state for endless calculations, the near-term solutions are all NISQ (Noisy Intermediate Scale Quantum) solutions. The "Noisey" term means the qubits have a finite quantum state and are unable to implement full quantum error correction. At Q2B, there was a lot of debate around the use and benefits of NISQ Computers; I certainly believe that there are uses and benefits of these machines.
There are many different types of quantum computers right now. Some current examples are Superconducting, Trapped Ion, Photonics, Cold Atom/Atom Array, and Annealing.
Superconducting Quantum solutions seems to be where many companies are making progress. The most well-known players are IBM, Google, Rigetti, and Intel. The systems operate in the RF spectrum and are cooled in dilution refrigerators to milli-Kelvin temperatures. These systems achieve large numbers of qubits (100+) and seem to have the fastest operation. The shortfall seems to be the gate fidelity, the duration a qubit maintains its state and is useable.
Trapped Ion Systems
Trapped Ion systems are being developed and are the focus of Ion-Q, Quantinuum (a new combined company of Honeywell's quantum group and Cambridge Quantum), and several others. These systems also achieve large numbers of qubits by using the stable electronic states of ions or charged atomic particles to create the qubit. Electromagnetic fields and lasers manipulate the ions for the states and motion of the qubits. The shortfall of this technology seems to be the speed at which calculations can be made, but the qubits have a much longer gate fidelity.
As you would expect, Photonic Quantum Computers are based on Silicon Photonics, integrated photonics circuits that create Qubits. The two leaders in this area are Xanadu and Psi Quantum. The consensus is that these systems have excellent gate speed but lack gate fidelity.
Cold Atom/Atom Array
Cold Atom technology is similar to Trapped Ion but uses neutral atoms for quantum computing. Some companies making systems based on this technology include Atom Computing, ColdQuanta, and QuEra. They have similar advantages and disadvantages to the trapped ion technology in that they have slower gate speeds but have somewhat high gate fidelity.
D-Wave, the leader in this technology, actually sold the first quantum computer in 2019 to Los Alamos National Lab. Annealing Quantum computers are a specialized version of a quantum computer leveraging the attributes of quantum physics to solve optimization problems and problems that can be solved using probabilistic sampling. The basic functionality of these systems is centered on the physics behind energy minimization. D-Wave has built 5,000 qubit machines but has been hampered by limitations in the problems it can solve. D-Wave is currently working to augment its machines with superconducting materials for gate-based systems similar to the aforementioned technologies.
Quantum Computing Software
While there are many variations in quantum computing hardware technologies, the software that needs to be developed to control and utilize these machines will be just as complex.
You probably have heard about the unsolvable problems quantum computers can solve. Molecular reactions, drug discovery, financial market analysis, complex climate analysis, and encryption and decryption are just some examples. The challenge with solving these problems is the need for a multi-disciplined community whereby the pharmaceutical, chemistry, or finance expert, the software developer, and the quantum physicist work together to generate a quantum algorithm for a given problem.
The many layers within the system add complexity: control and programming of the qubit and quantum machine, error control and correction, algorithms and libraries to model the problem the user is trying to solve, programming languages and APIs, and compilers. As you can see, there is probably just as much to develop in the software space as there is with the actual quantum computers.
Most quantum computers mentioned above are large, complex machines that users can connect to and program through the cloud. Many tools and software development kits are available today. Some are specific to the quantum computer solutions mentioned above, but solutions such as Amazon's Braket and QWare's Forge allow users to access multiple quantum hardware technologies.
The question remains whether NISQ Computers are useful or if users need fully fault-tolerant solutions. In my opinion, there is so much to learn in quantum computing and so many different technologies in play that NISQ computers are undoubtedly valuable technologies and systems. These systems will allow the software community to progress and mature, develop algorithms, and grow the overall technology.
Can Quantum Computers be used today? Absolutely. There are current applications like some machine learning solutions, optimization problems, and some crypto applications that can benefit from the current offerings in the short term. Activities and research in these areas will help mature all aspects of the quantucommunity's'squantucommunity's's hardware, software, and applications. As the technology ramps, so will the uses and benefits of quantum computing.
The quantum community is growing and wants to grow faster. Every company I talked to at this conference is looking to hire. Academia needs to help develop quantum experts in all areas and disciplines as this will be a growing need in the future. Those of us already in the industry need to partner with academia to foster the next generation of quantum computing talent.
understanding the unique chALLENGES
Benchmark works with supercomputing leaders, including in quantum computing, to develop the novel manufacturing processes required to bring unique quantum designs to reality. The unique challenges of quantum computing require unique materials, form factors, temperature tolerances, and other technologies thadon't't conform to well-established manufacturing methods. The engineers at Benchmark have supported emerging supercomputers for decadeswe'rere excited to support quantum computing as it continues to progress.
Ready to have a conversation about your quantum computing challenges? Contact us!