Quantum Processing Units (QPUs)

Courtesy: Shutterstock/Bartlomiej K Wroblewski

For additional references on Quantum Computing, see: A History of Quantum Computing, Quantum Generative Adversarial Networks, Quantum Artificial Intelligence, Hybrid Quantum-Classical Algorithms, and Quantum Computing in Finance.

While many are familiar with the above image of a quantum computer, it is not the only type of hardware that facilitates quantum computation. Quantum computing uses quantum-mechanical phenomena, such as superposition, entanglement, and interference. However, there are various ways that scientists can produce these quantum characteristics. Entanglement is when two or more particles become linked together so that their properties remain connected even when separated by large distances; this allows for faster communication between qubits (the basic unit of quantum computing). Superposition is when a particle exists in multiple states simultaneously, allowing for more complex calculations than traditional computers because each qubit can represent multiple values simultaneously. Interference is when two waves interact, creating either constructive or destructive effects, which can amplify certain signals while canceling out others. In the same way that classical bits can be represented as 1s and 0s, heads and tails, or electric voltage thresholds, qubits can also be produced in numerous ways, each with unique advantages and disadvantages. The following will touch on five of the most popular types of quantum computers:

• Superconducting Transmon Quantum Computers
• Photonic Quantum Computers
• Trapped Ion Quantum Computers
• Cold Atom Quantum Computers
• Topological Quantum Computers

While the field of Quantum Computing is still in its early days of development, there are several types of QPUs that have shown greater promise toward realizing a fault-tolerant system with sufficient coherence times and a scalable structure.

Superconducting transmon quantum computers

Superconducting transmon quantum computers use electrical circuits made from superconductors cooled to very low temperatures, near zero Kelvin, to achieve their functionality. The transmon qubit comprises two Josephson junctions connected in parallel, which are then coupled to an inductor and capacitor.

Figure 2.1: (a) Schematic drawing of a Josephson junction. (b) SEM micrograph of a Josephson junction fabricated in this work which is realized as two superconducting Al layers separated by a nm thick Al oxide layer. The actual junction area is marked with a green box. The duplicated structures are due to the shadow evaporation under different angles

The Josephson junctions create a potential barrier between the two sides, allowing for the creation of energy levels that can be manipulated by changing the voltage applied to them. By using microwave pulses to these energy levels, it is possible to control the state of the qubit and thus perform calculations on it. The downside of these systems is that they are prone to errors due to the fragile nature of the junctions, which require extremely low temperatures to operate effectively. This is necessary to mitigate the potential for disturbance from nearby atoms, magnetic frequencies, or various waveforms. Outside noise can disrupt the superposition and entanglement of the qubits in the quantum system making the calculations unreliable. Additionally, maintaining a supercooled system is neither economically sustainable nor environmentally sustainable, depending on the cooling mechanism. See AIs Carbon Footprint for a brief overview of the hidden cost of compute-intensive processes. While these downsides are meaningful, there are also significant advantages, such as the ability to easily connect multiple systems allowing for the QPU to scale to higher numbers of entangled qubits, a necessary component for realizing real-world applications of quantum computing algorithms.

Photonic Quantum Computers

Photonic quantum computers use photons to represent and manipulate information. Photons are particles of light that can be used to store and process data in a quantum state. Photons have certain properties that make them ideal for performing calculations quickly; they travel fast and can interact with each other easily when compared with electrons or ions used in other forms of QPUs. Photonic quantum computers work by using optical components such as mirrors, lenses, and waveguides to control the movement of photons. These components are arranged in an array known as an interferometer which splits a single photon into two separate paths and then recombines them again at the end.

https://photonlab.hajim.rochester.edu/research1.html

By manipulating these paths with different optical elements, it is possible to create various interference patterns that can be used to encode information into the photon’s state. This encoded information can then be manipulated with the quantum phenomena mentioned earlier. Additionally, because photons don’t interact with matter as strongly as electrons or ions do, they’re less prone to environmental factors such as temperature or humidity levels, allowing them to operate at room temperature. The disadvantages of photonic quantum computers are their scalability and complexity. They can also be relatively costly as they require a large number of optical components and lasers to operate effectively.

Trapped Ion Quantum Computers

Trapped ion quantum computers use ions, which are charged atoms, as qubits. Electric fields hold the ions in place, forming a potential well. This potential is modified using time-dependent electric fields, which create a wall preventing ions from escaping. The ions are cooled down so their motion is quantized, and photons are used to transition between the qubit’s ground and excited states. The qubits can be manipulated using laser frequencies, which can be arranged in a one-dimensional chain. Multi-qubit operations can be performed by carefully tuning the laser frequency, which causes the ions to recoil together. This phenomenon is called the Mossbauer effect.

Schematic diagram of a 5-qubit ion trap, a rudimentary quantum computer. The five calcium ions trapped by electrodes represent five quantum bits. Laser beams can perform quantum transformations on the ion strings and are also used to cool and stabilize the individual ions. https://www.researchgate.net/figure/Schematic-diagram-of-a-5-qubit-ion-trap-a-rudimentary-quantum-computer-The-five-calcium_fig3_220476032

The main advantage of trapped ion systems is that they do not require extremely low temperatures like superconducting transmon systems do. However, this comes at the cost of increased complexity since more lasers need to be used for manipulation purposes.

Cold Atom Quantum Computers

This artist’s illustration shows six finely tuned lasers being used to slow down atoms inside NASA’s Cold Atom Lab, which chills atoms to almost absolute zero. Credits: NASA/JPL-Caltech

Cold atom quantum computing takes advantage of atoms cooled to a low enough temperature so that their behavior becomes predictable enough for manipulation using lasers, similar to what is done with trapped ion systems mentioned above. However, unlike trapped ion systems where only one type of particle is manipulated (i.e., ions), cold atom systems can be functional with multiple types of particles, including neutral atoms, thus making them more versatile than trapped ion systems but also more complex due to the lack of uniform properties across various particles. This requires precise adjustments to the quantum system when seeking to manipulate different particles simultaneously within the same environment or in sequence. This quantum system has similar disadvantages to superconducting transmon quantum computers regarding the requirement for super cold operating temperatures.

Topological Quantum Computers

Lastly, there are topological quantum computers which are based on Majorana fermions where two entangled qubits form a single unit whose state cannot change without breaking the entanglement bond between the two qubits. This provides an additional layer of protection against external influences, mitigating the impacts of noise and decoherence, leading to better performance overall when compared to other forms of QPUs discussed previously. In topological quantum computers, a qubit is composed of a group of anyons. An anyon is a quasiparticle that obeys fractional statistics and exists in two-dimensional systems. It is neither a boson nor a fermion, and in addition to having fractional exchange statistics, they can exist in multiple quantum states simultaneously, which makes them attractive for use in quantum computing. Operations are performed by “braiding” the worldlines of the anyons, and measurements are taken by fusing the particles.

A topological quantum computation scheme, showing the braided worldlines of several anyons.

One of the benefits of using topological quantum computers is the ability to encode and store data for more extended periods due to the increased stability afforded by the particular architecture. Topological quantum computers are not entirely free from the impacts of noise and outside forces and require a high degree of control and accuracy, requiring specialized equipment and knowledgeable practitioners. Various institutions are pursuing topological quantum computing as a viable mechanism for scaling and mitigating errors associated with other quantum computing systems. Topological systems are still in their infancy, but recent progress indicates a promising future.

Conclusion

In conclusion, various Quantum Computers are already available, under development, or being researched as promising candidates for the next generation of computational processing units. These include Superconducting Transmon Quantum Computers utilizing electrical circuits cooled to near absolute zero temperatures, Photonic Quantum Computers utilizing photons and manipulation through laser pulses, Trapped Ion Quantum Computers using charged particles held inside electromagnetic traps, Cold Atom Quantum Computers taking advantage of the predictability of cooled atoms, and Topological Quantum Computers based on Majorana Fermions and the braiding of anyons. These systems allow scientists and practitioners to process information faster than ever before, making it possible to solve problems that were previously beyond the computational limits of classical computers. These technologies and approaches to Quantum Computing offer unique advantages and disadvantages. The optimal system may vary depending on the application requirements, similar to using CPUs, GPUs, and TPUs in classical computing. This gives users various options when deciding the best hardware to use when tackling a given problem.

For additional resources, visit www.quantumai.dev/resources

We encourage you to do your own research.

The information provided is intended solely for educational use and should not be considered professional advice. While we have taken every precaution to ensure that this article’s content is current and accurate, errors can occur.

The information in this article represents the views and opinions of the authors and does not necessarily represent the views or opinions of QuAIL Technologies Inc. If you have any questions or concerns, please visit quantumai.dev/contact.