The way that Nvidia co-founder and chief executive officer Jensen Huang sees it, his company’s role in the ever-emerging quantum computing field is no different than that in other industries where AI and accelerated computing play roles. Nvidia is an enabler and facilitator. It makes things happen.
“Nvidia doesn’t make quantum computers, but we dedicate ourselves to creating accelerated computing stacks to enable quantum computers,” Huang said during the GPU maker’s inaugural Quantum Day at the tail end of its recent week-long GPU Technical Conference in San Jose, California. “We do the same with self-driving cars. As you know, Nvidia’s probably more integrated into the world of automobiles and autonomous vehicles and we work with just everybody in some way to advance autonomous vehicles, and yet we don’t build cars.”
The same can be said about robotics – Nvidia has its hands in the industry, offering models and libraries, but it doesn’t build robots.
“We don’t build quantum computers, and yet we’re deeply integrated into the quantum computing industry and we create libraries,” Huang said. “CUDA-Q is the programming model for hybrid classical accelerated quantum. We have cuQuantum, libraries that help you simulate quantum circuits, and DGX-Quantum, to do error correction in quantum computers. We partner with them, we support them, we help them in any possible way.”
As he noted, Nvidia has been active in the quantum space for several years, partnering with a growing number of the myriad vendors in the industry and lending its AI and accelerated computing expertise for reference architectures, libraries, and tasks like designing the chips and systems.
Huang also got a lesson earlier this year about how volatile the industry can be, when he said during the Consumer Electronics Show in January that useful quantum systems were still probably 15 years to 30 years away, sending quantum stocks tumbling by as much as 60 percent.
The company’s first Quantum Day – there will be a second one next year – in part was a way to soothe any hard feelings about that slip and blip and to allow Huang a chance at mea culpa: “This is the first event in history where a company CEO invites all of the guests to explain why he was wrong.”
A Quantum Research Center
Quantum Day was not the quantum computing’s only place in GTC. Sprinkled among the news and discussions about advances in AI and Nvidia’s latest efforts, from expanded GPU lineups to product roadmaps to AI initiatives for the enterprise, were next steps in its quantum plans. Key among them is that Nvidia is planning to build a research center for building advanced technologies for quantum computing.
The Boston-based Nvidia Accelerated Quantum Research Center, or NVAQC – Huang said the company chose the city to be closer to research-focused universities like Harvard its Quantum Initiative in Science and Engineering unit and MIT’s Engineering Quantum Systems group – will integrate quantum hardware from the likes of Quantinuum, Quantum Machines, and QuEra with Nvidia-powered AI supercomputers to research and develop quantum processors and other technologies and address challenges like error correction in qubits.
The center will house a Nvidia GB200 NVL72 rackscale supercomputer that will include 576 Blackwell GPUs – that is four racks, or half of a SuperPOD – and will use the company’s Quantum-2 InfiniBand networking to lash the racks together. It will feature the DGX Quantum architecture and allow researchers to orchestrate what will be hybrid quantum-classical systems.
“As quantum computers continue to develop, they will integrate with AI supercomputers to form accelerated quantum supercomputers capable of solving some of the world’s hardest problems,” wrote Nicholas Harrigan, product marketing manager for quantum at Nvidia. “Integrating quantum processing units (QPUs) into AI supercomputers is key for developing new applications, helping unlock breakthroughs critical to running future quantum hardware and enabling developments in quantum error correction and device control.”
A Lot Of Partners
There also were a number of announcements in and around GTC about plans by quantum computing companies to leverage Nvidia technologies. Pasqal said it is integrating its neutral-atom qubits and cloud platform with the open CUDA-Q quantum-classical platform to give organizations more tools for developing and running quantum applications. Last week, Pascal announced it will build EuroQCS-Italy, a quantum computer that will be hosted in Italy. The French company was awarded a procurement contract from the European High Performance Computing Joint Undertaking (EuroHPC JU).
SEEQC said it created the first digital interface for connecting quantum chips to GPUs to allow for real-time error correction. The technology includes SEEQC’s Single Flux Quantum product integrating with CUDA-Q.
Quantum Rings integrated its quantum circuit simulation technology with CUDA-Q, QuEra worked with Nvidia to develop a transformer-based AI decoder for error correction through CUDA-Q and data from QuEra’s neutral-atom quantum chip, and Q-CTRL partnered with Nvidia and Quantum Circuits to reduce the compute costs that come with error suppression using its Fire Opal software.
Breakthroughs Are Being Made
Nvidia’s Quantum Day came against the backdrop of several advancements in addressing such critical challenges like error correction, which is difficult to achieve because of the fragility of qubits, which can easily be torn apart by light, vibration, and other disruptions. However, quantum heavyweights Google, Microsoft, and Amazon Web Services each outlined methods they developed that they say address the error correction hurdle.
In addition, researchers led by a professor at Queen Mary University of London suggest that superconducting quantum systems could run at room temperature, which is important because it was assumed that such computers would need to operate at very cold temperatures, which could limit their use.
Such advancements have helped roil the waters of the ongoing debate about the estimated time of arrival for useful and fault-tolerant quantum computing that Huang stepped into back in January. When announcing the company’s Majorana quantum chip in February, Microsoft executives said it could reduce the time to useful quantum computing from decades to years.
During a wide-ranging panel discussion with leaders of a number of quantum companies, Huang asked that question, with some at a high level noting that it could be a decade or more, but added a number of caveats, including the likelihood that quantum systems will work side-by-side with classical supercomputers, essentially taking on those workloads that the supercomputers can’t handle. Alan Baratz, CEO of D-Wave – which has its own conference this week and its own announcements – noted several times during the panel discussion that his company already has its annealing quantum computing on the market.
However, several people noted that applications are already running on some of their systems and the number of qubits they’re putting into them are growing.
Peter Chapman, executive chairman and former CEO of IonQ, said the industry still needs to mature, which will include embracing two or three modalities – which right now include superconducting, trapped ions, neutral atoms, photonic qubits, and topological qubits, among others – and going through mergers and other consolidation.
There is more commonality among the various quantum companies than might be expected and there is hope that there will more sharing and working together in the future to drive the development of quantum computing.
“New ways to build qubits are being found every day, but I think over the few years we will start to coalesce to probably two, may three different approaches and some of us will probably come together because we do share basically the underlying technology,” Chapman said, noting that it takes a long time to move from a startup to a company of Nvidia’s stature. “It’s fine for the quantum industry if it’s going to be another ten to fifteen years to get to where Nvidia and all the other giants are.”
That said, he added: “It’s just not going start then. It’s starting today.”
https://www.nextplatform.com/2025/04/01/nvidia-says-it-will-be-an-accelerator-of-quantum-computing/amp/
