Quantum computers have recently attracted a lot of attention for their ability to solve problems in a matter of hours, where the best supercomputers can estimate the age of the universe (i.e., tens of billions of years). Their real-life applications are diverse and range from drug and materials development to solving complex optimization problems. Therefore, it is mainly focused on scientific and industrial research.
Traditionally, a "quantum advantage" is required in terms of raw computing power: we want to compute (significantly) faster.
However, today's power consumption issues may also require investigation, as today's supercomputers sometimes consume as much electricity as small cities (which may actually limit the growth of their computing power). By 2020, information technology will account for 11% of global electricity consumption.
Why care about the energy consumption of a quantum computer?
Since a quantum computer can solve problems in a few hours that would take a supercomputer tens of billions of years to solve, we naturally expect it to consume less energy. However, building such a powerful quantum computer will require solving many scientific and technological challenges, perhaps a decade or more of research.
A simpler goal is to create a less powerful quantum computer that can perform calculations at the same time as a supercomputer but consumes less energy.
These potential energy advantages of quantum computing have already been discussed. Google's Sycamore quantum processor consumes 26 kilowatts of electricity, far less than a supercomputer, and runs quantum testing algorithms in seconds. After the experiment, scientists proposed a classical algorithm to simulate the quantum algorithm. The first proposal of the classical algorithm required more power, demonstrating the energy advantages of quantum computing, but other more energy-efficient proposals soon followed.
Therefore, the question of power benefits is still open and an open topic for research, especially since Sycamore's quantum algorithm has yet to identify a "useful" application.
Superposition: A Fragile Phenomenon of Quantum Computing
To find out whether a quantum computer provides an energy advantage, we need to understand the basic laws by which it works.
Quantum computers manipulate physical systems called qubits to perform calculations. A qubit can take two values: 0 ("ground state", for minimum energy) and 1 ("excited state", for maximum energy). It can also hold a "superposition" of 0's and 1's. How we interpret superposition is still a hot topic of philosophical debate, but in simple terms it means that a qubit "can be both in the 0 state and at will 1 with some" "probability amplitude".
With this possibility, we can greatly simplify the working principles of a quantum computer by saying that it implements an algorithm that performs arithmetic operations "at once" on many numbers (in this case, 0 and 1 at the same time). This advantage becomes apparent as the number of bits increases: 300 stacked qubits can simultaneously represent 2 to the power of 300. For example, this is roughly the number of atoms in the observable universe, so it remains simple to map multiple states simultaneously on a supercomputer. unreal.
However, the fundamentals of quantum theory tell us that if the amplitude value of this probability were "measured" by another physical system, the superposition would be destroyed: the qubits would shrink to a value of 1 or 0, causing an error in the calculation.
A striking example of such annihilation is the absorption of a photon (a light particle with a small energy packet) by a qubit. If it does, it's because it's not in its maximum energy state (where it can absorb energy, the energy of photons). Thus, the photon and through it the "neighborhood" of the qubit indirectly "finds" the value of the amplitude, destroying the superposition. This is called decoherence.
In general, the challenge is to make sure that the qubits are sufficiently isolated to prevent information loss: we cannot allow photons or other particles to interfere with our qubits. This is difficult because even qubits must be controlled: they cannot be completely isolated.
This lack of protection is a major source of error in qubit-based calculations. For example, one of the more mature qubit technologies experiences an error every 1000 operations. If you consider that a typical quantum algorithm takes 10¹³ operations, you can see that there are a lot of them.
Maintaining obstacles requires energy
The computing power costs of quantum computers will primarily be driven by the need to "protect quantum data". For example, to avoid the above problem, the qubit environment must be set close to absolute zero (-273 °C) to ensure that the environment is not filled with photons. This is a process that requires energy.
Some other technologies, such as quantum error correction, also store quantum information and can improve operational accuracy. However, apart from the problems they cause, these technologies also lead to very high energy costs, as additional error detection algorithms or error detection qubits etc. are required.
In short, the more precisely we want to perform operations on a qubit, the more we protect it and the more energy we have to spend on it. In quantum computing, there is a very strong relationship between "error rate" and "energy". Therefore, understanding these correlations may enable the development of energy-efficient computers.
Is it possible to obtain quantum energy?
Some theoretical studies have been able to estimate the energy costs of building quantum computers, but in non-optimized systems, often using simplified computer models without exploiting the correlation between error rate and energy in particular.
Exploiting this correlation can lead to powerful optimizations that reduce the energy consumption of the algorithm. In practice, this requires an interdisciplinary approach that includes understanding the fundamental phenomena that cause decoherence, modeling quantum error correction algorithms and codes, and all the "engineering" required to control qubits. Thus, it is possible to calculate the minimum energy costs required to solve various problems, focusing on the error probability of the algorithm considered "acceptable".
As we can see, for qubits of excellent quality (that is, qualities that are practically unattainable today), there are tasks for which a quantum computer can use a hundred times less energy for the computation time than the best modern supercomputers (in a comparative sense, both can complete the task in a reasonable time will know). This power factor of 100 is also an indicator: more energy savings can be imagined with more improvements.
This is because quantum computers perform calculations using fundamentally different processes from quantum computers: one deals with qubits and the other deals with qubits. Therefore, the number of transactions for the same activity and even with the same processing time can be drastically different. Furthermore, the processes performed on a quantum computer will involve fundamentally different physical processes than those performed on a supercomputer. Taken together, these two observations show that, conceptually, even with the same computation time, a quantum logic process consumes more energy than a conventional logic process, but fewer quantum logic operations could mean that a quantum computer would ultimately be more powerful. effective.
Of course, this example sometimes comes from theoretical calculations based on very optimistic assumptions. However, this suggests that one of the main advantages of quantum computing may be active before computation.
This article is republished by The Conversation under a Creative Commons license. Read the original article. 
Quote : Could energy efficiency remain the biggest selling point of quantum computers? (2022, October 26) Retrieved October 27, 2022 from https://phys.org/news/2022-10-energy-efficiency-quantum-greatest-st Strength.html.
This document is subject to copyright. No part may be reproduced without written permission, except in commercial fairness for personal study or research purposes. The content is provided for informational purposes only.