There was a time, not that long ago, when artificial intelligence operated in a quiet corner of the tech world without a lot of attention or interest. Quantum computing is in a similar position today. Could it become the next revolution in technological change?

This was the possibility on the table as a host of quantum industry chief executives and researchers spoke in a series of panel discussions during “Quantum Day” at the Nvidia GTC conference in San Jose on Thursday. The event was put together in the aftermath of comments made by Nvidia CEO Jensen Huang during the Consumer Electronics Show in January: He declared that it would be “15 years for useful quantum computers and that would be on the early side. Thirty years is probably on the late side. If you picked 20 years, a whole bunch of us would believe it.”

The remarks triggered a massive sell-off in quantum company stocks on Wall Street and prompted a frustrated Alan Baratz, CEO of D-Wave Quantum Inc., to fire back in a blog post and later stated, “When he says something stupid, people don’t think it’s stupid.”

In his introductory remarks at the GTC Quantum Day gathering, Huang offered his own “mea culpa” of sorts for the stir his remarks caused. “This is the first event in history where a company CEO invites all of the guests to explain why he was wrong,” Huang told GTC attendees. “We care deeply about this ecosystem. This whole event is going to be like a therapy session for me.”

Tackling the quantum challenge

The reality, as explained by a parade of panelists, is that quantum computing is difficult, complex and laborious work. Companies are also finding that the puzzle of how to make quantum useful for real-world applications can be solved in any number of ways.

The premise behind quantum is that it differs from classical computing systems which rely on “bits,” a one or a zero, for computational processing. Quantum bits, or qubits, can be ones and zeros, which enable them to crunch numbers at impressive scale, significantly outperforming traditional computers. The qubit is a tiny particle that carries computer information.

Building computers that can use qubits effectively has been an enormous challenge. D-Wave is using a process called “annealing” that applies an external magnetic field to manipulate qubits. Recently D-Wave announced that a scientific paper had validated its ability to solve a problem impossible for most classical computers. “There are still hard computational problems that are out of the range of classical and AI isn’t going to solve that either,” Baratz said.

Other companies are leveraging different ways to corral quantum power. Quantinuum Inc., which is majority-owned by Honeywell International Inc., and IonQ Inc. use trapped ions or charged atoms as qubits. That enables the firms to link them together as a base circuit and carry out computer operations.

A few companies have extended this concept even further by placing quantum computers on a chip. Seeqc Inc., which has Nvidia as an investor, is developing a digital quantum computing platform that will leverage chips to power all the core functions of a quantum computer. “Unless we can figure out how to build quantum computers on a chip, we’re never going to get there,” said Chief Executive John Levy.

This notion of a QPU or quantum processing unit offers the tantalizing possibility that it can ultimately be combined with a GPU or CPU to dramatically enable accelerated computing. Nvidia’s contribution in this area has been CUDA-Q, an open platform for programming hybrid quantum systems.

In November, Nvidia announced that it would be linking its CUDA-Q platform to supercomputing sites in Germany, Japan and Poland to power QPUs inside high-performance systems that also leverage GPUs and CPUs.

“It isn’t whether one will replace the other,” said Peter Chapman, president and CEO of IonQ. “They are working together today. I expect it will be a QPU, GPU and CPU all working together.”

Working to reduce errors

A key problem confronting the industry is that quantum computers are notoriously error-prone. Finding a way to solve this, known as error correction, has emerged as the “holy grail” for the quantum industry today.

In a recent paper, the quantum research team from Google DeepMind outlined how it has been using AI algorithms to introduce a decoder system that can accurately identify computing errors. The solution, AlphaQubit, is a neural-network-based decoder that can be trained on hundreds of consistency checks and spot when a qubit begins to behave incorrectly.

Error correction will likely be a key factor in determining which of the varied approaches to quantum computing ultimately succeeds, according to Simone Severini, general manager of quantum technologies at Amazon Web Services Inc.

“[Error correction] is going to be the theme that will determine which modalities emerge in time,” Severini told the GTC audience. “We believe that error correction is going to be really important for quantum computers to deliver on their promise.”

Scaling up for use cases

What is the promise of quantum computing? The consensus opinion, as expressed by multiple speakers at the GTC event, was that quantum’s vast computational power will create its own tornado of use cases. Scientific research and drug discoveries were the two most frequently mentioned, along with a host of applications that demand massive computing power.

For now, the competitive landscape around quantum computing has boiled down to a race for building machines with the most qubits. Atom Computing Inc. has developed a neural atom quantum computer that is populated with 1,180 qubits, one of the highest benchmarks in quantum capacity.

However, as noted by Ben Bloom, founder and CEO of Atom Computing, the industry must boost qubit capacity significantly higher for quantum to have a major impact. “These things that are going to change the world? You have to get to a million qubits,” Bloom said.

Meanwhile, technology investors continue to pour money into the quantum ecosystem. According to data supplied by Crunchbase, quantum computing startups raised $1.5 billion in venture funding during 2024 alone, nearly doubling the $768 million raised the previous year.

One speaker on Thursday described quantum’s progression as a volume knob that gets turned up steadily over time. When the industry reaches a point where the noise level is impossible to miss is anyone’s guess, but there was no shortage of optimism about quantum’s future at GTC this week.

“It takes time,” said AWS’ Severini. “Quantum computers are going to be so impactful that it’s going to be a great party in the end.”

Photo: Mark Albertson/SiliconANGLE