IBM makes major strides in quantum with Volume and AI

• IBM Quantum Volume creates a new way to assess a quantum computer’s capabilities as a whole system, rather than just based on its number of qubits.
• IBM unveils research into the intersection between quantum computing and machine learning.

IBM, ever present in the development of cutting-edge technology, is a leader in the quantum computing space and brought its IBM Q System One to market in January. This came after the company provided access to three of its quantum systems through the cloud. TBR notes that providing access to actual quantum systems, and not simply quantum simulators via the cloud, differentiates IBM’s capabilities from those of peers. Currently, a key purpose of providing access to these quantum systems through the cloud is education. IBM is learning from past challenges by getting out ahead of innovation and making internal and external learning and development a priority. This will ensure that trained internal personnel are in place once IBM Q achieves commercial application and research viability as well as help accelerate the rise of IBM Q by educating those outside IBM who might be involved in developing technological advancements, at least at the algorithmic level.

Quantum Volume assesses quantum capability of a whole quantum system

IBM unveiled Quantum Volume, which is a way to measure the capabilities of an entire quantum system, including qubits, software and overall functionality. IBM aptly demonstrated the value in measuring the functionality of a quantum system by more than just the number of qubits. Qubits are very complex, and factors including the quality of the qubit and the impact of unwelcome external stimuli need to be assessed as qubits are highly sensitive to environmental influences.

Aspects such as gates, connectivity, algorithm errors and compilers are all assessed by Quantum Volume. The capability of the computer is then determined, and it is categorized with a number, which serves as a rating of sorts. This rating is a key to the entire quantum computing market, TBR believes, because it provides a relatively unbiased way to measure the capability of a quantum computer. The current process, in which the number of qubits is used to measure capability, omits essential factors.

More significantly, IBM contends it has discovered a metric for quantum advancements comparable to Moore’s Law for classical computing advancements. Given quantum computing will be adopted essentially algorithm by algorithm, this new metric could help guide the broader user community about when a given algorithm will be ready to move to quantum computing based on the advancements in the technology and the complexities of the algorithm in question. This information can then be used to optimize the deployment layer based on mapping the algorithm to the known performance of the different qubits within the systems.

IBM evaluates the parallels between quantum computing and machine learning

IBM published some of its research, in which it evaluated the applications of quantum computing in conjunction with AI and machine learning to address additional emerging demands. Further details regarding this research can be found in an article published in Nature titled, Supervised learning with quantum—enhanced feature spaces. At its core, AI is simply the evaluation and analysis of massive data volumes that help train a system. In classical computing this process can be time-consuming, but when digital transformation is added to market offerings such as connected cars, time is of the essence and cannot be squandered performing these types of tasks. Quantum computing would not only reduce time to insight but also improve the accuracy of the insights gathered.

Quantum computers can analyze and evaluate data much faster than a classical computer and can also process more complex data sets. These increased levels of speed and complexity will enable machine learning to be more insightful and, therefore, more applicable to more complicated use cases. Key use cases that IBM highlighted for these capabilities initially are model training, pattern recognition and fraud detection. These use cases are fundamental to a connected world, where bad actors would have the potential to cause great harm if they tamper with systems, as would machine malfunctions. If data can be evaluated and extrapolated into real-time applications faster, the potential dangers of an increasingly connected world can be mitigated more quickly.

The market implications of quantum developments are vast and will be rapid

Although individually these announcements may seem small in the scheme of quantum computing, when combined with IBM’s existing breakthroughs in the technology, they demonstrate the breadth of the market IBM’s quantum capabilities will be able to impact once quantum advantage is achieved. Quantum Volume enables IBM to determine use-case efficiencies for quantum computers once commercial availability is attainable. The ability to combine the capabilities of quantum computing with AI will accelerate digital transformation dramatically and launch society into the next technological revolution much faster, as more rapid time to insight will open new avenues of exploration.