Skip to Content

The Rise of Quantum Computing

Machine Learning has always been aided and, at the same time, limited by computer power. As a result, research and development are laser-focused on quantum computing and how it can offer the computing power to do things that classical computing cannot. In this chapter, we’ll contemplate how the rise of quantum computing has influenced and will continue to support Machine Learning innovations.

The Rise of Quantum Computing

Content Summary

How Quantum Computing Evolved
Quantum Machine Learning
Industry Applications
Quantum Computing for ITOps
Limitations
Conclusion

How Quantum Computing Evolved

To this point, we’ve been talking about ML in the context of classical computing. With their thousands of classical CPU and GPU cores, the biggest classical computers, supercomputers can solve many complex problems. However, the truth is, even supercomputers can’t manage to hold the number of combinations involved in realworld issues—supercomputers don’t have enough memory. Since the 1980s, scientists like physicists Richard Feynman and David Deutsch have been exploring how quantum mechanics can improve computing speed and create solutions where none currently exist.

Quantum computing uses the collective properties of quantum states (superposition, interference, and entanglement) to execute computations. Quantum computers are being designed to solve particular computational problems, like integer factorization, significantly quicker than classical computers.

Large tech companies, including notably IBM, Microsoft, Google, and AWS spend a lot of R&D money developing analog quantum computers. IBM has built a network of partners in its IBM Quantum initiative. It includes a worldwide network of Fortune 500 companies, academic institutions, researchers, educators, and enthusiasts committed to driving innovation in this area. IBM, Microsoft, and AWS all offer cloud-based quantum computing options, primarily quantum computing simulators, so that interested parties can experiment with algorithms, test quantum hardware, design software, and explore viable applications for the technology as it matures.

Quantum Machine Learning

As Gartner pointed out in its Hype Cycle for Data Science and Machine Learning, 2021, quantum ML could enable a subset of ML algorithms to be run using the quantum computing paradigm. However, Gartner notes that only a limited number of quantum ML algorithms currently exist: “We have yet to see any evidence that ML could benefit from quantum computing over traditional alternatives. However, the parallel nature of some ML techniques could make quantum computing a viable path to explore. Increasing awareness of quantum ML capabilities plays a key role in determining its potential value.”

Industry Applications

As we mentioned earlier in this chapter, Google, IBM, Microsoft, and other major tech companies put substantial resources toward quantum computing research as they attempt to pioneer breakthroughs for industries like medicine, supply chains, financial services, and so on. From banking and financial services to healthcare, general science, and cybersecurity, quantum machine learning offers businesses opportunities to improve services and expand exponentially.

Financial Services

Banks and finance firms, for instance, are already using machine learning techniques such as reinforcement learning for algorithmic trading, and they use Natural Language Processing for risk assessment, financial forecasting and accounting and auditing. Some practical uses of quantum computing in the near future include the following:

Algorithmic Trading: Trade decisions can be sped up and complex models can be simplified by using quantum reinforcement learning methods in algorithmic trading.

Asset Pricing: Banks have been using Recurrent Neural Networks (RNNs) to run time series predictions. Institutions like JPMorgan are considering using them for asset pricing models but RNNs consume a lot of computing power. However, it can be advantageous to use parameterized quantum circuits and quantum Long Short Term Memory units that allow users to predict outcomes based on evolving processes from historical data.

Fraud Detection: Currently a wide area of concern that is being tackled by classical computing ML, fraud detection can benefit from quantum clustering algorithms when used to detect anomalies.

Volatility Prediction: Quantum methods can also be applied to tracking and predicting changes in a security’s price. A density matrix can be produced by deep quantum neural networks, and the implied volatility of an option can then be calculated using its corresponding element in that matrix.

Cybersecurity

Although quantum ML will improve anomaly detection, data protection is another cybersecurity area to benefit. Modern cryptography and its algorithmically encoded data have risks. Most notoriously, hackers can intercept the cryptography key to decipher the data or use powerful computers to predict the key. With enough compute power, a hacker will win.

Researchers are cautioning companies that more secure cryptography keys will be needed in the future. One solution is to make cryptography keys totally random and impossible to solve mathematically. Randomness is fundamental to quantum behavior, making quantum ML perfect for creating cryptography keys that are impossible to reverse-engineer.

Healthcare and Pharmaceutical Advancements

In previous chapters, we discussed how AI is being used in the early detection of diseases, treatment, and research. But there are also applications in the discovery of new drugs. The procedure currently relies on molecular simulation, which involves modeling how particles interact inside a molecule to configure molecules capable of treating a particular disease. Sounds pretty complex, right? And it requires a lot of computational power.

Quantum computers could speed up this process, whittling a typical 10-year cycle for bringing a new medicine to market down considerably. Pharmaceutical giant, Roche, is already working with Cambridge Quantum Computing to design and implement noisy-intermediate-scale-quantum algorithms for early-stage drug discovery and development.

Quantum Computing for ITOps

In the future, ITOps solutions may employ cloud quantum computing power to move beyond dependence on vendors and experts and design and implement self-learning and self-healing IT operations. The power of quantum computers combined with advanced AI models can help an IT system constantly learn about its environment and improve itself while adapting to changes.

ITOps teams may be able to let autonomic computing take on the complexities of network monitoring and maintenance and even have the system itself build a constantly expanding knowledge base that surpasses what humans can reasonably provide.

Limitations

Quantum computing is promising, but it’s still fairly limited in actual application, including ITOps teams and the organizations they support. Generally, systems can only scale to tens of qubits (quantum bits). This means that algorithms executed on these systems are mainly exploratory. Although quantum computing can theoretically bring dramatic benefits to some classes of data, a big challenge is encoding. For quantum ML to scale successfully, great quantities of data must be encoded and loaded into the quantum system.

Likewise, new algorithms that can take advantage of capabilities offered by near-term noisy quantum systems are in the discovery stage. Research in this evolving field continues, and some scientists question the potential applicability of quantum computing in ML. And, of course, hardware and software must evolve as well.

Today, research and development tend to focus on creating different quantum algorithms for ML kernels. Companies like IBM have prototyped ML algorithms for select use cases, and IBM specifically has developed a roadmap to reach 1000+ qubits by the end of 2023.

Although quantum ML is still in an early stage, Gartner has noted that “potential applications of quantum computing in artificial intelligence and ML include quantum search, recommendation algorithms, quantum algorithms for game theory, and quantum algorithms for decisions and learning.”

Conclusion

In this chapter, we looked at the exciting area of quantum computing and how it may affect industries—and the ITOps that support them—in the future. IT organizations can start experimenting with cloud-based offerings like IBM Quantum, Azure Quantum, and Google Quantum AI. In our next chapter, we’ll find out how the trend towards hyper-personalization built on data, analytics, and ML impacts IT organizations and benefits companies.

    Ads Blocker Image Powered by Code Help Pro

    Ads Blocker Detected!!!

    This site depends on revenue from ad impressions to survive. If you find this site valuable, please consider disabling your ad blocker.