<< back Page 2 of 3 next >>

Quantum Computing Gets Real But IS Not Ready for Prime Time


Ultimately, quantum computing provides the computational power needed for the growing surge of big data requirements, Kothari said. “Quantum computing is the ‘need of the hour,’ due to the increasing volume of complex data which is getting stored every day in petabytes. I see the role of quantum increasing in many work streams, including fraud analytics, marketing analytics using unstructured data, and so on. That’s because quantum not only brings better computational power to data scientists but also the power of quantum algorithms such as Grover’s algorithm and entanglement algorithms—tools which bring a unique flavor with respect to unsupervised learning.”

The power and speed of quantum computing “cannot even be comparable to what traditional computers are able to do today,” said Joe Kim, executive VP of engineering and global CTO at SolarWinds. “You are changing 1’s and 0’s to qubits—making speed irrelevant as a comparable feature to what can be done today.” This will be instrumental in tackling huge data and AI problems, he continued.

Quantum computing will benefit data environments in other ways as well. For example, it is “uniquely efficient at solving not just molecular modeling but also instrumental aspects of artificial intelligence and machine learning,” said Halliday. “The ability to run instruction sets in a massively parallel environment is par-amount and, when  coupled with quantum-resilient security data, it will be more secure against theft.”

Quantum computing also offers a more powerful option for decision making. This is most apparent when solving optimiza-tion challenges, in which the goal is to find the best choice out of many possible options, said Garrison.“These exist in virtually every industry and tend to be difficult to solve for.” Examples include finding the most cost-effective route for shipping goods, determining the most efficient way to extract resources from a mine, and seeking the most productive resource allocation involved in a production line, said Garrison. “Quantum computing offers a distinct advantage in that it provides a much speedier answer. This has the potential for significant impact, as time saved can translate into cost savings and revenue generation.”

However, some industry experts feel quantum computing may be overkill for many data-related applications, as it may be too costly, complicated, and, ultimately, unnecessary. “While quantum computing is posed to revolutionize several application domains, such as financial modeling, cybersecurity, and drug development, it is not a panacea,” cautioned Ion Stoica, co-founder of Anyscale, and also co-founder of Databricks. “For example, quantum computing is not built for ingesting and processing large amounts of data, for implementing business logic (which requires an auditable trail), or simulating complex industrial models, such as robots moving on a factory floor plan. For these applications, conventional computing is still the best solution for the foreseeable future.” 

"In scientific research, the ability to achieve seemingly impossibly high levels of parallelism will accelerate research that is bottlenecked on computation.  Protein folding and other microbiological simulations may lead to more miracle drugs or personalised medicine.," said Guy Harrison. He expects the situation to resemble the 1970s and 1980s where Cray and others offered "supercomputers" that were capable of orders of magnitude greater computation. However, just as in the 1970s and 1980s, he said, most applications won’t need massively faster computation, when their bottleneck is already on network and database bandwidth.


While quantum computing holds great promise for enterprises, there are risks that need to be addressed along the way. Data security is a top concern. “Technologies like quantum services and blockchain are promising when it comes to compelling use cases and futuristic business applications, but they haven’t been vetted thoroughly for data security issues,” Kothari cautions. “For quantum to become mainstream and accepted at the enterprise level, cloud and quantum vendors must collaborate with cyber and data security experts to ensure security is holistically embedded throughout the offering.”

Another challenge is that, as information and insights move through enterprises at blazing quantum speeds, it may be more difficult to vet and assure the trustworthiness of results. Quantum may be employed “to solve problems that are so difficult that there’s no real way to confirm if the response is accurate or not,” said Garrison. “There could have been mistakes in the architecting of the mathematical algorithms, or undetected hardware issues. Building up trust in the responses provided by quantum hardware will be a different journey for each company. For some—in healthcare and pharmaceuticals, for example—it will be important to have mechanisms to confirm the results. In other industries it might be safer to reach further with quantum computing but have risk mitigation processes in place to protect against mistakes that may accompany the aggressive use of the technology.”

<< back Page 2 of 3 next >>


Subscribe to Big Data Quarterly E-Edition