Google Achieved Technical Quantum Supremacy. Competitive Quantum Advantage Is Yet To Come.

With today’s publication of Quantum Supremacy Using A Programmable Superconducting Processor in the science journal Nature Google has hit a critical milestone. In short, the company has reached what is known in the field as technical “quantum supremacy”, using a quantum computer to solve a problem that could not have been solved with a classical computer. As University of Texas at Austin Professor Scott Aaronson has aptly observed, this is the quantum computing equivalent of the Wright Brothers’ flight at Kitty Hawk. Just as that first 12-second flight proved the technical feasibility of air travel, Google’s achievement marks the dawning of the quantum computing age.

I, like many others, was not sure this day would arrive. I first became fascinated with the potential of quantum computing after reading a Scientific American article about Shor’s algorithm in 1994. But I was somewhat skeptical that we could harness individual qubits in a way that would yield reliable computing systems. The ingenuity of researchers in the field—both within and outside of Google—have proven me wrong.

Google’s achievement today paves the way for a future era of “quantum advantage” when the technology can fuel breakthroughs in areas like drug discovery, aerospace design, or network optimization that would be impossible using today’s classical computers.

Google’s Breakthrough—And What Must Come Next 

The experiment highlighted in Google’s peer-reviewed publication today is what is known (in industry jargon) as a “sampling problem”. Google’s 53-qubit computer Sycamore was asked to calculate the possibility of each possible outcome occurring across a random circuit. Sycamore’s algorithm took just 200 seconds to complete the task.

How much faster is that than what could be achieved by today’s most powerful supercomputers? Google argues it would take 10,000 years for a supercomputer to do the same calculation. IBM has pushed back, arguing it would take the Oak Ridge National Laboratory’s Summit supercomputer just 2.5 days to complete the task.  While that may be true, the fact is that adding just 1 qubit to the problem and the computer would more than double the lead over Summit as IBM reckons it would then take 8 days. As IBM did for the classical algorithm, further optimizing the Sycamore algorithm could extend that lead even further. This spirited back and forth about the differential between Sycamore and Summit is interesting certainly. What is not debatable, however, is the fact that Google has achieved technical quantum supremacy.

That is a major advance—but hardly the endgame. This was, after all, not a practical application but a very specific laboratory problem suited to a very particular hardware. That means there is still much work to be done before quantum supremacy - the technological superiority of quantum computers “in a lab” - translates into quantum advantage—the creation of real business value for industry players.

That will require major progress in the number of qubits manipulated and in their quality (with current technology, they are unstable and therefore prone to computational errors). At the same time, the industry should focus its efforts not only on the hardware, but also on developing and improving associated algorithms. Making both more efficient (hence requiring less qubits) and more tolerant to noise will allow us to capture value even as the hardware is being perfected.

From Supremacy to Advantage

So how will the quantum computing market evolve in the years ahead? Based on extensive BCG research, we expect some key technical issues with quantum computing could be solved—and the first pockets of value created—within five years. This will be the so-called “Noisy Intermediate-Scale Quantum (NISQ) era”, with devices that are very powerful in specific use cases, albeit with approximate results due to noise. These computers will provide speed-up over classical computers in areas like material simulation, which will in turn reduce expensive and time-consuming trial-and-error lab testing. Total value created during this period is expected to range from $2 billion to $5 billion annually.

In 10 years’ time, we expect quantum computing hardware will have both much more qubits and the ability to self-correct, allowing for near-real-time simulation for activities such as risk assessment by financial services firms. This is the “broad quantum advantage” period, which we project will create $25 billion to $50 billion in value annually.

Finally, in 20 years or so, we expect quantum computing will have fully arrived. Full-scale fault-tolerant quantum machines will bring scale and stability to their computing abilities. They will be true game changers in a number of industries. Biopharmaceutical companies, for example, should be able to use quantum computing to design de novo biologics that have minimal side effects. The potential is huge: a projected $450 billion to $850 billion in value annually.

The First Flight of Quantum Computing

When the Wright Brothers landed among sand dunes in Kitty Hawk, they knew regular passenger flights would not happen overnight. Yet, they paved the way for more sophisticated machines, eventually transforming aviation into a massive, global industry and changing the very nature of travel. Google hasn’t solved all the challenges facing quantum computing. But they have shown all of us that the technology has lift-off.