Holy Grail of Science
And that's the kind of passion that's going to be around quantum computations right now, because whoever's the first one to be able to develop a sustainable quantum computer is going to rip off a big pile. Cryptography and data hacking, looking for new materials and drugs, optimization, risk assessment -- the solution to these problems is almost instantaneous. Many people are scary, and it's not for nothing: all our data and savings are going to be in jeopardy. This could lead to a world disaster that's already been called a quantum apocalypse.
But as long as the power of quantum computers is too small, the devices can only show the ability to solve the problems, not the ability to solve them. To hack a cryptographic algorithm with an open key, RSA needs about 20 million cubes. Cubbit is a quantum analog of a bit that can take not only 0 and 1 but both of these positions at the same time. In the science world, it's called superposition, and that's what makes it easier to find an answer, but it's not easy to create it.
For example, Google plans to release a quantum computer with 1 million cubic metres only in 2029, and today's quantum computers have a maximum of 100 cubic metres. The more they are, the more difficult it is to connect them directly. In addition, modern cubes are unstable and eventually lose their quantum state, and the computation results contain a large number of errors. These factors greatly delay the creation of a quantum computer.
Quantum benefits here and now
Mathematics and physics have learned to apply the benefits of quantum computing on a classical computer without waiting for the Q day. Quantum-inspired algorithms find solutions within an acceptable time frame. The best modern algorithms already find them 95-99% close to optimal. The devices help calculate the optimal route for spacecraft, and they develop new drugs and materials.
The location of wind power stations, oil wells and ambulances is also optimized by quantum-inspired algorithms, and this can be done with any schedule, such as trains or vacations. Of course, sometimes a small company's leave plan can be done manually. But if there are too many variables and constraints, it is not possible to do such work with a simple overrun: it will take decades to find an answer. The way out before the quantum-inspired algorithms were invented was by one -- deliberately not including a few indicators. Of course, it affected the quality and effectiveness of the solution.
To use the algorithm, you need a special software-hardware complex. It's called a decisionr or a solver. It's used to solve optimization problems when you have to look at millions of combinations. Solver allows you to take all the constraints into account, and its work does not depend on the subject area for which the task is performed. The decisionr works in a single logical scenario using a vector of variables and a matrix of limitations.
Solvers are divided into two types: the first is specialized, which includes the Yandex. Marshrutization platform. It solves the problems of logistics and choice of the optimum route, taking into account traffic, traffic lights and road maintenance. The second is industrial or universal solvers, such as Fixstars or IBM CPLEX. They solve optimisation problems with a large number of variables and limitations. They are used to find new materials or to make a production schedule. Universal Solver is the advanced optimization technology.
The Need for Solvers
Until Day Q, when real quantum computers are available, mankind is still far away. But companies from different fields and countries are already applying and introducing quantum technologies. No wonder optimization is needed wherever there is planning and consistency, because it helps to save budgets, resources and time.