Trending
Prev Story
Next Story

Nvidia teams up with IBM and Fermilab on a quantum codesign platform

The Next Web (TNW) •
November 16, 2026 • 04:31 PM
• By Darius Popa
6
6 reads 0 reactions

‘Scaling quantum and classical together will open new frontiers,’ IBM’s Jerry Chow says of the quantum codesign platform with Nvidia and Fermilab.

1 / 6
Swipe Photos

From left: an IBM Quantum System Two, an Nvidia DGX server and cryogenic hardware photographed by Fermilab.

Nvidia said on Monday that it plans to work with IBM and the US Department of Energy’s Fermi National Accelerator Laboratory on a quantum codesign platform. The three will use it to research modular quantum computers, error correction and scientific applications, Nvidia said in its announcement .

At the core of the platform, Nvidia DGX B200 systems will sit beside IBM quantum computers, including machines that run IBM’s Nighthawk processor . Nvidia called Nighthawk the fastest quantum processor IBM has delivered. IBM and Fermilab published their own announcements on the same day.

Researchers will use the platform to build and test error-correction codes, decoding algorithms and applications that mix quantum and classical computing. The applications target high-energy and nuclear physics. The work will also study how classical systems can grow alongside several linked quantum processors.

Fermilab will steer the work towards error correction and physics. The aim is to help set the requirements for future fault-tolerant quantum computers, according to Nvidia.

“Scaling quantum and classical together will open new frontiers in high-energy and nuclear physics and beyond,” said Jerry Chow, IBM Fellow and chief technology officer of quantum-centric computing.

Anna Grassellino, Fermilab’s chief technology officer and director of its SQMS Center, called it a “proposed platform” in the statement. None of the three partners gave a start date or a cost.

Quantum processors make errors, and those errors limit how long and how complex their calculations can be. Error correction spreads each piece of quantum information across several physical qubits. A classical program called a decoder then reads the measurements and works out which errors most likely happened.

Nvidia also released a new decoder, TCBP-JOBS , in its CUDA-Q QEC 0.9.0 library. Nvidia said it decodes up to 13,000 times faster than the leading existing algorithms, which it did not name. In research on new quantum computer designs, that cuts the computing time from 12 weeks to 45 minutes, the company said.

Fermilab has used Nvidia’s quantum software before. In September, it said Nvidia’s CUDA-Q Logical layer cut a design job from five months to three weeks. In October, Nvidia simulated a 104-qubit fault-tolerant quantum computer on Germany’s JUPITER supercomputer.

The announcement covered two more projects. IBM and Japan’s RIKEN institute combined an IBM quantum computer with RIKEN’s ROQUO supercomputer, which runs on Nvidia GB200 systems. They calculated binding energies for two protein and drug-molecule pairs, a problem that matters in drug discovery.

Their models held up to 12,635 atoms. The previous benchmark for this kind of hybrid calculation was 303 atoms. The work is a finalist for the 2026 ACM Gordon Bell Prize. Japan’s NEDO, an agency of the trade and industry ministry, supports it.

Nvidia is also adding a quantum interface to Slurm, the open-source software that schedules jobs on supercomputers. IBM is contributing a connection through its Quantum Resource Management Interface. Centres would then book quantum processors next to CPUs and GPUs with the commands they already use. Today, each centre builds its own custom setup, Nvidia said. It did not say when the Slurm features will be ready.

Get the most important tech news in your inbox each week.

Prev Swipe Story Next

Same Story from Other Publications

Compare how 6 other independent news outlets reported on this story.

6 Publications
TechCrunch
33m ago
Trump says ex-Nvidia CEO Jensen Huang's affair with an alleged Chinese spy did not threaten national security | TechCrunch

Trump says ex-Nvidia CEO Jensen Huang's affair with an alleged Chinese spy did not threaten national security | TechCrunch

Trump contradicted his own government's intelligence officials, who reportedly concluded that Alice Huang is a Chinese spy.

CNBC World
48m ago
Market's Nvidia obsession obscuring an opportunity in this name

Market's Nvidia obsession obscuring an opportunity in this name

Investors might be overlooking Disney earnings.

NPR News
55m ago
U.S. tightens controls on remote access to AI technology amid Nvidia fallout

U.S. tightens controls on remote access to AI technology amid Nvidia fallout

The new rule closes a loophole that existed that allowed Chinese companies to still access advanced AI chips remotely, despite U.S. restrictions on their physical sale to China.

The Verge
1h ago
The Age
1h ago
ASX set to fall as Nvidia drags Wall Street lower

ASX set to fall as Nvidia drags Wall Street lower

More drops for stocks in the artificial-intelligence industry are weighing on the US market, led by Nvidia.

Sydney Morning Herald (SMH)
1h ago
Trump teases Kim Jong-un meeting as he heads to Asia, backs Huang for Nvidia return

Trump teases Kim Jong-un meeting as he heads to Asia, backs Huang for Nvidia return

The US president continued to cast doubt on whether Alice Huang was actually a Chinese spy and said former Nvidia boss Jensen Huang had “learned a big lesson”.

Related Stories

Technology
Build #2612