Nvidia has unveiled a system for linking quantum computers to its AI chips. The technology could significantly speed data processing and open new avenues in medicine and materials science, Bloomberg reports.
For “particle-level” computing to move into practical use, it must be integrated with conventional hardware. Nvidia has built dedicated software—NVQLink.
The software is intended to help design a new generation of supercomputers and reduce qubit error rates.
“It not only fixes problems within today’s qubit counts but is also aimed at the future,” said the company’s CEO Jensen Huang at the Nvidia GTC event in Washington.
He plans to scale quantum computers from today’s hundreds of qubits to tens and hundreds of thousands. To that end, the company has enlisted 17 firms specialising in such computing.
Huang added that the firm will work with the US Department of Energy to develop seven new supercomputers. They will be equipped with AI chips and sited at federal research centres.
The high-performance systems will partly help the US maintain and develop its nuclear arsenal. They will also be used to research alternative energy sources such as nuclear fusion.
The largest of the supercomputers for the Department of Energy will be built with Oracle and equipped with 100,000 Blackwell chips.
In October, Google began developing artificial intelligence to manage the process of thermonuclear fusion, teaming up with Commonwealth Fusion Systems.
A supercomputer for new drugs
Eli Lilly announced a partnership with Nvidia to build a supercomputer to search for new medicines and speed their development.
Using it, the firm will train AI models on the basis of millions of experiments testing potential drugs.
“Lilly is moving from using artificial intelligence as a tool to embracing it as a scientific partner,” said Eli Lilly’s senior vice president and chief AI officer Thomas Fuchs.
Some models will be available on Lilly TuneLab—a platform that allows biotech companies to access AI-based tools.
The world under Nvidia’s sway
Speaking in Washington, Huang said the United States will prevail in the AI contest if the entire world, including China, runs on his company’s systems.
The CEO of the American chipmaker praised President Donald Trump for expanding investment in manufacturing, but warned of the risks of worsening relations with China.
“Support for energy growth from the state completely changed the situation. If that had not happened, we might have found ourselves in a bad situation, and I want to thank President Trump for that,” Huang said.
He added that Nvidia has orders for Blackwell and Rubin chips worth $500bn, with deliveries scheduled five quarters out. Even so, the company still needs the Chinese market.
Good news for Trump
Huang plans to make a joint announcement with South Korean companies that “will please Trump and South Korea”.
The Nvidia CEO will meet world leaders and top executives of South Korean firms at the Asia-Pacific Economic Cooperation CEOs summit.
“Samsung, SK, Hyundai, LG, Naver—if you look at the whole Korean ecosystem, these companies are my close friends and very good partners,” he told reporters.
Partnerships
Nvidia also set out new details of a partnership with Nokia. It will buy 2.9% of the Finnish telecoms giant for $1bn.
The chip designer additionally introduced a new product line called Arc, intended for networking equipment.
Nvidia intends to work with Nokia to improve the energy efficiency of 6G base stations.
“We are going to use this new technology and will be able to modernise millions of cellular nodes around the world,” Huang said.
The Finnish company will issue more than 166m shares. The funds will be directed to AI projects and other corporate tasks.
Nvidia also announced a partnership with Palantir Technologies, which works closely with the US government. The chip designer will help accelerate solutions to logistics problems.
Earlier the company made a series of strategic investments: $5bn in Intel, $500m in Wayve and $667m in Nscale. It also said it intends to invest $100bn in OpenAI.
Autonomous vehicles
Nvidia announced a new technological platform for self-driving cars—Hyperion. The company is working with Uber to create a robotaxi network.
“This will be a new computing platform for us. I expect it will be quite successful,” Huang said.
Meanwhile, Lucid Group announced plans to launch a mid-size electric vehicle with Level 4 autonomy using Nvidia computing systems.
The transport of the future will require minimal human intervention and approach fully automated Level 5 systems.
The company will use a combination of sensors, including cameras, radar and lidar.
The electric-vehicle maker also plans to use Nvidia’s software tools at its factories to improve quality and speed up production.
In September, Amazon entered the US robotaxi race—the first public launch of Zoox took place on the Las Vegas Strip.