In November, IBM introduced its 127-qubit quantum processor called Eagle, showing that the dream of achieving true quantum computing power seems to be on track -- though work must continue before real-world applications of the technology can be realized.
Demand for the development of quantum computers, which apply quantum physics for computations and data storage, stems from the belief they will solve problems far more accurately, faster, and at less cost than current compute machines. Many players, from AT&T and Amazon to Zapata Computing and Xanadu, have their hands in some aspect of quantum compute development.
Breaking the 100-qubit threshold for quantum processors is a next step in attaining so-called “quantum advantage.” That is when quantum computers should show they can vastly outstrip classical computers. IBM says Eagle is their proof that quantum advantage will be attainable by 2023.
There can be frothy chatter with fuzzy numbers and promises thrown around that may oversell quantum computing. “Some of the hype out there, some of the qubits you hear about don’t actually work,” says Robert Sutor, chief quantum exponent with IBM. He explains that producing individual qubits is not enough; they must interconnect and function together to count. “If they can’t, we can’t execute the quantum computing model to do the sorts of calculations that we want.”
The 100-qubit threshold, Sutor says, was a significant roadblock in the efforts to attain quantum advantage. A processor with several hundred qubits, he says, all working together may be needed to fully demonstrate the exponential improvement quantum computers should represent over classical computers.
In addition to strides being made in hardware, quantum computing is finding a place in the cloud. “Microsoft and Amazon both came up with this idea of getting third-party hardware access to their environments,” says Konstantinos Karagiannis, associate director of quantum computing services with business consultancy Protiviti. He also expects Google to do something similar soon. “They are actually working on their own machine.”
Other activity in this space includes the October launch of the AWS Center for Quantum Computing in partnership with the California Institute of Technology, a joint effort at a facility in Pasadena to build quantum computers. Investments are being made into use cases for the financial world, Karagiannis says, to at least have a proof of concept in the works. “This is going to be a tough thing to play catchup on,” he says. “We’ve seen this in machine learning.”
You Snooze, You Lose
Companies that waited around on machine learning, Karagiannis says, may have found themselves lagging while their peers pursued such resources. “Trying to make a splash in machine learning now is very difficult,” he says. “All of the good stuff has already been established.” Something similar may happen with quantum computing.
Hiring personnel who can support innovation in quantum computing, Karagiannis says, can be difficult because it may require more than a quick update to traditional development skills to work in this space. “You have to have some understanding of linear algebra, some basic understanding of the physics behind it,” he says. “Machine learning background helps, too. Finding that synergy of talent is a little challenging.” Karagiannis expects there to be 30% to 40% increases annually in the need for such talent.
For now, it seems the more technology developers who contribute to the development of quantum computing the merrier. “We need a lot more machines and we need them to be a lot better,” Karagiannis says. As more companies discuss ways to create quantum processors, he says they can bring new methods to figuring out best practices in production. “We need to continuously encourage these materials, science, research, and approaches.”
Not every business or sector will necessarily see an immediate need for quantum computers, Sutor says. Financial services, chemistry, and logistics may benefit from this next evolution of computers, he says.
The pace of progress in developing quantum computers is determined largely by the scale of interconnected qubits, Sutor says. Quality of qubits is another factor, he says, which includes keeping them cold and in the dark, away from interference. This is to cut down on background “noise” that might otherwise interfere with processing capabilities.
With their ever-escalating power, quantum computers should make tasks that were once monumental much simpler to complete. For example, Sutor says a chemistry calculation conducted two years ago that took 4.5 billion quantum calculations required some 290 days to complete with a classical computer. “We can do that in seven hours now,” he says. “We have changed the way classical and quantum computers work together. We’ve improved the algorithms; we’ve improved the quantum computers themselves.”
Sutor says the next step in 2022 for IBM will be to produce a 433-qubit processor.