Nuclear Power

[phkrause]

US aims to create nuclear fusion facility within 10 years, Energy chief Granholm says

VIENNA (AP) — The Biden administration hopes to create a commercial nuclear fusion facility within 10 years as part of the nation’s transition to clean energy, U.S. Energy Secretary Jennifer Granholm said Monday.

https://apnews.com/article/granholm-nuclear-fusion-energy-emissions-416f497b842ce94292ec6fb009a4087a?

[phkrause]

Powerful plans

If everything comes together, Virginia will be home to the world’s first grid-scale nuclear fusion power plant. The goal is for the facility to harness the futuristic clean power and generate electricity from it by the early 2030s.

[phkrause]

From years to weeks: How a fusion startup aims to use AI to speed breakthroughs

A trio of companies including Nvidia announced Tuesday that it's using AI to accelerate the development of fusion that's eluded innovators for decades.

https://www.axios.com/2026/01/06/nvidia-siemens-fusion-energy-startup-cfs?

[phkrause]

Fusion within reach

Illustration: Sarah Grillo/Axios

Fusion could be commercially possible within a few years — thanks to AI, Axios energy correspondent Amy Harder writes. Why it matters: Fusion has eluded scientists for decades. 💬 "I had my first congressional hearing on fusion 45 years ago, and the experts on the panel said it's 50 years out. So I'm getting excited," quipped former Vice President Al Gore, now a leading environmentalist and cleantech investor, in an interview with Axios. Fusion is the atomic process that powers the sun. Its potential sounds too good to be true: a clean, near-limitless energy source. The big picture: AI is poised to help fusion in two ways, according to Andrew Holland, CEO of the Fusion Industry Association: Machine learning is helping companies build virtual models that let them test and optimize designs without ever building anything. Partnering with corporate heavyweights Nvidia and Siemens, well-funded fusion startup Commonwealth Fusion Systems is employing this exact strategy. Taming fusion plasmas — glowing clouds of charged particles that are inherently unstable and prone to disruptions. Disruptions happen in milliseconds — too fast for human operators. "Machines can move much faster than humans, but they are also able to see patterns in the data as well," Holland said. Google has been at the forefront of this research, alongside Commonwealth Fusion Systems. 🔎 What we're watching: Most fusion startups are aiming for commercial deployment by the early 2030s. And if AI solves fusion, it will help power the future of AI itself.

[phkrause]

⚡ Scoop: OpenAI eyes fusion deal

Photo illustration: Lindsey Bailey/Axios. Photo: Chip Somodevilla/Getty Images

OpenAI is in advanced talks to buy electricity from Sam Altman-backed fusion startup Helion Energy, Axios AI+ co-author Ina Fried writes. Why it matters: The move shows OpenAI is serious about tapping fusion energy to help meet the ChatGPT-maker's massive power demands. Fusion — a commercially elusive process that powers the sun and stars — could be key to providing the AI industry with abundant, clean energy to meet its growing needs. Helion believes it's on the cusp of "scientific breakeven" — a key step to reaching the point where the company's fusion process can generate more energy than it consumes, a source told Axios. 🔬 Between the lines: OpenAI CEO Sam Altman, an investor in Helion, has stepped down as the board chair, the source said. Keep reading.

[phkrause]

CERN hails test on transporting antimatter as a scientific success

Scientists at CERN took some antiprotons out for a spin in a never-tried-before test drive. Tuesday's experiment was hailed as a success. About 100 antiprotons traveled in a truck around CERN's Geneva campus and were transported back into the lab. Each particle had a corresponding antiparticle, like an antiproton, exactly matching the particle but oppositely charged. If the opposites came into contact, they would “annihilate” each other and set off energy. The mass in the test was tiny, so the worst that could happen was losing the antiprotons. Read more.

RELATED COVERAGE ➤

WATCH: Nuclear research agency CERN runs delicate test

[phkrause]

☢️ Microsoft, Nvidia make nuclear push

Illustration: Rebecca Zisser/Axios Microsoft and Nvidia are joining forces to accelerate nuclear power buildup, Axios' Chuck McCutcheon reports from Houston. The initiative is the latest example of the tech industry leaning on nuclear's emerging potential to handle AI's huge energy needs. 🤖 Microsoft President Brad Smith unveiled the "AI for nuclear" initiative yesterday at the CERAWeek energy industry conference. Smith said the companies have created a solution "that hopefully will play an important role in expanding the construction of nuclear power." 🏗️ Darryl Willis, Microsoft's vice president for worldwide energy and resources industry, wrote in a blog post that AI tools can help identify documentation inconsistencies, unify plant construction data, and support "digital twins" — virtual replicas that allow engineers to test changes. Go deeper.

[phkrause]

⚛️ The race to build new nuclear reactors

Illustration: Brendan Lynch/Axios. Stock: Getty Images HOUSTON — As nuclear energy deals pile up, executives face a huge challenge: How to build multiple plants quickly, Axios' Chuck McCutcheon writes. Why it matters: Nuclear is seen as critical to powering AI data centers, but the supply chains, workers and permits needed to build plants haven't kept pace. 🔮 At last week's CERAWeek energy conference, industry officials predicted that companies' willingness to try new approaches — particularly with AI — will speed things up. Nvidia and Microsoft are joining forces to use AI tools to streamline nuclear construction, using virtual replicas that let engineers test changes. Aalo Atomics says it cut permitting time by 92% using Microsoft's Generative AI for Permitting solution, saving about $80 million a year. The goal is a "copy and paste" approach, said Aalo's Jon Guidroz. "We need to build this [nuclear] stuff the way server racks and data centers get built." Getting a qualified workforce fully up to speed to build many plants will take time, said Ross Ridenoure, Hadron Energy's chief nuclear officer. "There will be, I think, a shortage initially, until the training programs catch up with the demand," he told Axios. The bottom line: Hyperscalers are bringing a much-needed sense of urgency, said TerraPower president and CEO Chris Levesque. Keep reading.