This article is about how quantum computing technology deployment, and when the industry benefited article, experts its associated data through interviews with quantum computing by the more authoritative in the field of McKinsey & Company, the relevant research data collated. Here only for sharing, quantum finishing off for industry decision makers.
This paper describes the advantages and the basic principles of quantum computing, and is illustrated by the data, how the future of quantum computing industry and enterprise deployments will use a quantum computer, final note, at this stage, facing quantum computing rapid advances in technology, which companies To this end it should be made ready.
- Enzymes, pharmaceutical companies darling
- Determining molecular modeling structure, difficult
- Break out the time-consuming
Quantum computing as a new calculation method, it will be able to solve the problem within a few hours past the classic computer impossible, to find science and social distress for many years the answer, and for a variety of companies to release unprecedented potential ability. These are calculated to achieve and quantum altered properties.
- Rome was not built in preparation
However, the speed of today's technological advances, can no longer be directly applied to past experience speculation. Quantum computing potential so great that every business leader should have a basic understanding of the technology works, know that such technology can help solve any kind of problem, and when the potential of using this technology, their own needs What kind of preparation do?
Compared with a variety of computing today conducted on laptops, workstations and supercomputers, quantum computing is a completely different from the previous calculation method. First of all, be clear, it does not replace the equipment, but by using the basic principles of quantum mechanics, can solve those for the current computer is difficult, concrete, typical, or even no solution is very complex statistical properties The problem.
Classical Quantum Computation and calculation and comparison
- Fundamentally different: bits and qubits
- Different results: the uniqueness and the possibility of
And a quantum computer, due to the natural physical properties of the qubit, which has exponentially workspace. Since a large number of quantum computers parallel paths simultaneously, which makes quantum computers in solving special problems, may speed becomes fast. They can provide a variety of possible outcomes in a small range, compared to the traditional single computer results, it can quickly make your answer faster approach. This allows the possibility to solve the problem of greatly increased.
- Quantum - Classic joint: a hybrid approach
It is this difference in the quantum computer has the potential to be much faster than today's servers and supercomputers. Quantum computers can simultaneously input a plurality of various calculations. And today's computers can only process a set of input and a calculation. For example, using a number of quantum bits (assuming of N), quantum computers can simultaneously enter up to 2 ^ N is calculated.
Before when quantum computing, as above described, although very clear, but when you dive into the actual working principles of quantum computer, you will find that potential in a quantum computer, still faces many challenges.
Technical barriers
Physical instability qubit
Quantum computing on the road, some of which are technical obstacles. For example, instability of the qubit. With a classical computer bit different, it may represent a combination of quantum bits 0 and 1 of the superposition state (superposition state), but the drawback is that, in the physical environment, which is superimposed state is not stable. Therefore, to ensure a place on a quantum computer chip does not interfere with any other position on the chip, it would need to do a lot of auxiliary work.
And more importantly, we need an interaction between qubits, this interaction also affect the physical stability of the qubit. Precisely because the presence of these interactions, a quantum computer could makes it possible to parallel computing.
Such control hardware interaction is very complex. Instability qubit will cause the input information is lost or replaced, thereby reducing the accuracy of the results, even give unreliable results. And to create a valuable, large-scale quantum computer would require hundreds of thousands or millions of qubits. Therefore, the current one of the few quantum computers can not handle so many qubits.
Solve math problems
From startups to research institutions, to Google, IBM and Microsoft like the software and hardware companies are working hard to overcome these obstacles. Including research and bears little resemblance to our algorithm currently used, today's black box looks completely different hardware, as well as contribute to the existing data into quantum bits supported software. But they still have a long way to go. Although quantum computing as a concept has existed in the early 1980s, but until September 2019, Google just announced that its Quantum computers just 2 minutes and 30 seconds to solve the traditional supercomputers require time-consuming to solve the problem of 10,000 years quantum computer was first proved by the fact that classical computers can handle it for a very complex issue, but it is more of a pure math, rather than anything that can be applied to the application (the commercial nature of the business operations also solve mathematical problems but to solve practical problems of operation is considered feasible applications) - to prove the advantages of quantum problems in real life, and of no practical use.
Range, but not the answer
The nature of quantum mechanics led to underestimate the advantages of its exponential growth rate. Today's computer to run in a very simple way: they use an algorithm to deal with a limited set of data, and provide a single result for you. The more complex quantum computer, after the data units into a plurality of qubits, qubits to operate so that the interaction between the qubits, which can be computationally intensive. This is also a quantum computer faster than a classical computer an overwhelming place. However, quantum computers can not provide a 100% accurate answer, the fact that reduced its value. In fact, the user is likely to get an answer within a small range. They might find themselves essentially been counted more than once in order to further narrow the scope of the quantum computer, and this result is a significant reduction in its multiple operational speed gain.
The results obtained by quantum computers is a range rather than a single definitive answer, which makes quantum computers do not sound as accurate as today's computers. However, this view is correct for the limited range of calculations, this is also one reason why a quantum computer would not replace the existing system. However, for solving a wide variety of extremely complex issues, the classic computer will spend a lot of time, quantum computers will eliminate the possibility of a lot of these problems, which can save a lot of time.
Quantum computer has four basic functions, they can be separated from classical computer zone:
- Quantum analog (Quantum simulation), a quantum computer simulation modeling of complex molecules;
- Optimization (Optimization), that is at an unprecedented rate to solve multivariable problems;
- Quantum Artificial Intelligence (AI), has a better algorithm, we can change the pharmaceutical and automotive industries such as machine learning;
- Prime factorization (The prime factorization), which could revolutionize the conventional encryption technology.
Chemical shorten development time by simulation
Seeking to develop new drugs and materials scientists often need to know the precise molecular structure to determine its characteristics and how it interacts with other molecules. Unfortunately, even relatively accurately model is difficult to smaller molecules using classical computers, because each atom interacting with other atoms in a complex fashion. Current computer almost impossible to accurately simulate even smaller number of atoms in the basic molecule, e.g., to a protein composed of thousands of atoms. That's why today's scientists are forced to use synthetic chemistry methods to physically measure its properties. Furthermore, the molecules generally can not be expected to work, the need for synthesis and testing of many, therefore, each optimization cycle is both expensive and time-consuming.The unique quantum computer, its physical nature is very suitable for solving this problem, because the intramolecular interaction atom itself is a quantum system. And experts believe that, in fact, a quantum computer can even be on the human body is the most complex molecular modeling. So every little progress in this direction will be to promote the faster development of new drugs and other products, and may offer a new treatment transformative.
Subversive speed, to solve optimization problems
In every industry, many complex business problems involve many variables. For example, companies should be placed in which position the robot factory floor? How to choose the delivery vehicles to ensure the shortest driving route? What is the most effective way to deploy cars, motorcycles and scooters to create a transport network to meet user needs are? How to optimize performance and risk finance portfolio? As above, a number of these cases only a few issues business leaders face in.
If classical computing to solve the above problems, it will be very difficult. In order to improve performance or loss independent drive input, it must be strictly limited in a number of variables which can be calculated in any variation. Therefore, any number of variables in the calculation of the variable must be strictly limited. Therefore, the calculation must be a classic and a complex calculation, which is a costly, time-consuming process.
Since the quantum computer can process multiple variables simultaneously, so that they can be significantly reduced in the range of possible answers within a very short time. Thus, the classical calculate the exact result can be obtained within a small range. Nevertheless, compared with quantum computing, its efficiency is still very slow. Because quantum technology eliminates many possibilities, so this hybrid approach will significantly shorten the time needed to find the best solution.
Quantum Artificial Intelligence accelerate the development of autopilot
Quantum computers may accelerate autonomous vehicles era of fast arrival. Among the many start-ups, Ford, General Motors, Volkswagen and other car manufacturers as well as new mobile field, the engineers through complex neural networks, for a long time to process video, images and lidar data. Their goal: to make use of artificial intelligence church car driving key strategies, such as when to turn? Where acceleration or deceleration? And how to avoid other vehicles and pedestrians? In this way, a series of intensive training artificial intelligence algorithms require calculations, with the increase of data, and an increase in more complex relationships between variables, making calculations become increasingly difficult. Such training needs may make the world's fastest computer work continuously for several days or even months.
Since the quantum computer can perform complex calculations of the plurality of variables simultaneously, they can be exponentially accelerated training of such artificial intelligence system. However, this is not achieved in the short term. Because the current, converting the data set into the classic quantum data remains a difficult task, so early quantum artificial intelligence algorithms made only negligible results.
Change network security concept
Quantum computing poses a serious threat to almost every company relies network security system. Most of today's online account passwords and the security of transactions and communications are protected by encryption algorithms such as RSA or SSL / TLS and the like. These systems enable companies to easily create can be shared by an authorized user data, while protecting it from outside attacks. For today's computers, we want to break the encryption requires enormous computing power. It is almost impossible with current computer fast enough to solve mathematical problems behind encryption of the actual structure. (The mathematical problem called the prime factor decomposition, because the encryption is built around the processing of large prime numbers) Thus, when data is stolen, usually due to poor implementation of network security protocols.
Since the quantum computer can perform multiple calculations simultaneously, they may destroy any classic encryption system. In fact, already there is a quantum algorithm to do this (Shor algorithm). Fortunately, there is no quantum computer capable of processing required to perform Shor's algorithm hundreds of thousands to millions of qubits. As previously mentioned, the current quantum computer can handle only tens of qubits. But between 10-20 years from now, this situation may change, then, will require a new round of quantum encryption technology to protect our most basic online services. Scientists and forward-thinking strategists have begun this revolution in quantum cryptography, trying to prepare for this turning point.
(Source: McKinsey)
Most companies do not have a quantum computer. Instead, they will receive a service through the cloud of quantum
Quantum computing is a complex technology. It is not an application today, tomorrow, millions of people appeared to be accepted. After talking with the rapid development of quantum ecosystem of dozens of experts, McKinsey & Company has the technology in the coming decades will be how to develop with a clear vision.
Quantum computer would be a few key organizations, institutions to develop and operate, expensive machine. Like Google and IBM hopes in a similar way Moore's Law, doubling the annual performance of a quantum computer. Promising force plus a small portion of startups, they will steadily increase the number of quantum bits which the computer can process. As this technology is new, its progress may be relatively slow. We estimate that, by 2030, only 2,000 to 5,000 quantum computer in operation. Because quantum computing challenges involve many parts, so to solve the most complex problems required hardware and software possible at the earliest in 2035.
Nevertheless, before this quantum computing will create value for some businesses. In the long run, companies will now depend on the quantum of service from their suppliers via the cloud. Amazon Web Services, Microsoft Azure and other companies have released a quantum goods and services. These products may cloud the rapid expansion of applications and needs.
在2022年至2026年之间，我们预计许多存在优化问题的企业将采用混合方法（量子+经典），其中部分问题将由经典计算处理，而部分问题将由量子计算机处理。在同一时间范围内，量子计算机可能变得足够强大，可以开始为化学、材料和制药公司处理有意义的分子结构模拟。而量子AI的到来还遥遥无期，我们预计量子计算机最早要到2020年代末期才能具备足以进行质因子分解的功能。
这一技术发展的时间表告诉我们，什么时候不同的行业可能从量子计算中获益最多。我们与之交谈的专家预计，到2025年，先进行业、全球能源和材料、金融以及（较小程度的）旅行和物流领域的先驱可能会开始从量子产品中产生巨大价值。 鉴于解决最复杂的医学问题需要模拟深度复杂的分子，药物的巨大回报可能要到下一个十年才会到来。
如下表所示，到2030年代中期，许多行业将有潜力从量子计算中创造重大价值。
（来源：麦肯锡）[/caption]
显然，为重大技术进步做准备是任何高管人员在进行投资组合时的关键思考。对于量子计算来说尤其如此，因为它具有极大的破坏性。通过求解经典计算无法实现的难题，量子技术可以使各种当前隐式知识变得明确。这不仅会彻底改变工艺流程，还会从根本上改变不同行业的劳动力。
（来源：麦肯锡）
大多数企业不会拥有量子计算机，他们将通过云获得量子服务
例如，在化学品和药物领域，当今的合成化学家必须创建实际的分子或固体，以测试有关潜在新药或新材料的假设。然而，这些物质通常无法按预期方式发挥作用，从而导致进一步循环的成本高昂且耗时的合成和测试。如果量子计算机能够按预期更快地对此类物质进行指数建模，那么公司很可能只需要很少的合成化学家。因此，这种数学上的确定性会取代其他行业的专业人士的专业知识和判断力，例如面临多变量问题的金融、保险、运输等行业。
Despite decades we are less likely to feel the social impact, but almost every industry prescient business leaders now should develop some kind of quantum strategy. Preparations factor depends, if you can benefit from this technology become the first wave of the industry, whether you have a business use case is mapped to the fledgling quantum advantage, and whether you believe you can get transformative or just increment of income.
The first wave of quantum computing industry
We believe, finance, tourism, logistics, energy and materials, and advanced global industry and other sectors will start to get significant value from mixed classical / quantum processes in the early 2020s. The first wave of these areas of business leaders need to quickly develop quantum strategy, otherwise, Barclays (Barclays), BASF (BASF), BMW (BMW), Dow Chemical (Dow), Exxon Mobil (ExxonMobil), etc. These have take a quantum computing strategy of innovative companies will be left behind. These leaders should consider how their businesses take advantage of the emerging quantum infrastructure.
Some businesses may now want to enter the labor market, and employing quantum developers to set up an internal team to study the system for urgent problems created algorithms. However, due to the quantum personnel currently in short supply, research universities are less likely to develop a short period of time sufficient quantum of top engineers to meet the rapidly growing demand. However, some companies may think that work directly with the development of quantum technology companies, which can meet their needs. But we are in the early stages of a long process to adapt to business needs, so the company still has the potential to meet the specific needs of its way to influence this development.
Protecting data assets, long-term archiving
In addition to these industries in the first wave of participating companies, there is another group of people should actively monitor the progress of quantum computing. According to Louisiana State University professor Jonathan Dowling saying, "If you have a need for confidentiality 10-50 years of trade secrets, you now need to start to worry." Whose business depends on decades of data security and confidentiality of the company, must vigilant on network security issues caused by quantum computing. The theme should be the top priority agenda of the chief information officer, business leaders must be confident that their company developed a plan to transition from the current cryptography to quantum cryptography security.
Protection of Trade Secrets
Technology is also your key technical experts should be monitored even if your business does not belong to these two types of quantum computing. Quantum technologies capable of not only a marginal improvement iterative technique, but also has the potential to change and subversive. This powerful technology will appear at an unpredictable rate, and the unpredictable effects. If you have trade secrets you do not want to be found by others, then, as a leader you should start now to prepare for the quantum computing.
About the Author
AlexandreMénard is a senior partner in McKinsey's Paris office, Ivan Ostojic is a partner of Zurich office, Mark Patel is a senior partner in the San Francisco office, Daniel Volz consultant Frankfurt office. The authors wish to thank Maximilian Charlet, Anna Heid and Lorenzo Pautasso contribution to development made herein, and MiklósDietz, Mathis Friesdorf, Eric Hazan, Nicolaus Henke, Anika Pflanzer and Henning Soller contribution.
About McKinsey
McKinsey & Company is a world leading global management consulting firm, was created in the United States by the University of Chicago Business School Professor James McKinsey (James O'McKinsey) in 1926. McKinsey adopt a "one company" partnership system, there are more than 80 offices in 44 countries, a total of more than 7,000 consultants. McKinsey Most customers are excellent large national companies, as ranked in the top "Fortune" magazine's top 500 companies.
Author | AlexandreMénard editor | Sakura
Disclaimer: this article for the purpose of high-quality information, the source tagging errors or if infringement, please hold of proof of ownership to contact us, we will promptly correct, delete all images whose rights belong to the organization cited here only quote, original articles reprinted need authorization.
