Microsoft is hiring a nuclear energy expert to help power its AI and cloud data centers
Microsoft CEO Satya Nadella speaks at an event on the company’s campus in Redmond, Washington, on Feb. 7, 2023.
Chona Kasinger | Bloomberg | Getty Images
Artificial intelligence takes a lot of compute power, and Microsoft is putting together a road map for powering that computation with small nuclear reactors.
That’s according to a job description Microsoft posted Thursday seeking a nuclear technology expert to lead the company’s technical assessment for integrating small modular nuclear reactors and microreactors “to power the datacenters that the Microsoft Cloud and AI reside on,” the job posting reads.
Specifically, Microsoft is looking to hire a “principal program manager for nuclear technology” and that person “will be responsible for maturing and implementing a global Small Modular Reactor (SMR) and microreactor energy strategy,” the job posting reads. Microsoft is looking to generate energy with nuclear fission, which is when an atom splits and releases energy that as a result of that splitting.
In January, Microsoft announced a multiyear, multibillion-dollar investment in OpenAI, maker of viral AI chatbot ChatGPT. Bill Gates, Microsoft’s co-founder, is also the chairman of the board of TerraPower, a nuclear innovation company in the process of developing and scaling small modular reactor designs.
TerraPower “does not currently have any agreements to sell reactors to Microsoft,” a spokesperson told CNBC.
However, Microsoft has publicly committed to pursuing nuclear energy from an innovator in the fusion space.
In May, Microsoft announced it signed a power purchase agreement with Helion, a nuclear fusion startup, to buy electricity from it in 2028. Sam Altman, CEO of OpenAI, is an early and significant investor in Helion.
Nuclear fusion occurs when two smaller atomic nuclei smash together to form a heavier atom and release tremendous quantities of energy in the process. It is the way in which the sun makes power. Fusion has not yet been recreated at scale here on earth, but many venture-backed startups are working to make it a reality due to the potential promise of virtually unlimited clean energy.
Interest in nuclear energy has increased alongside concerns about climate change in recent years, as nuclear reactors generate electricity without releasing virtually any carbon dioxide emissions.
The existing fleet of nuclear reactors in the United States were largely built between 1970 and 1990, and currently generate about 18% of the total electricity in the United States, according to the U.S. Energy Information Administration. Nuclear energy also makes up 47% of America’s carbon-free electricity in 2022, according to the U.S. Department of Energy.
Much of the hope for the next generation of nuclear reactor technology in the United States is pinned on smaller nuclear reactors, which Microsoft’s job posting indicates the company is interested in using to power its data centers.
Small nuclear reactors are, as the name suggests, smaller than conventional reactors and cheaper and quicker to build, since they’re designed with modular construction and not every piece of the reactor needs a bespoke manufacturing process. One primary criticism of the nuclear industry is that building reactors has come to require excessive time and money.
Charles Liang, CEO of Super Micro, speaks at the HumanX AI conference at in Las Vegas on March 10, 2025.
Big Event Media | HumanX Conference | Getty Images
Super Micro issued disappointing guidance on Tuesday, a week after the server maker provided preliminary results for the latest quarter that fell far shy of Wall Street’s expectations. The stock slid about 4% in extended trading.
Here’s what the company reported in comparison with LSEG consensus:
- Earnings per share: 31 cents adjusted vs. 50 cents expected
- Revenue: $4.60 billion vs. $5.42 billion expected
While the latest numbers were below analysts’ estimates, they were in line with early results that Super Micro disclosed last week. The company said at the time that revenue in the fiscal third quarter would be between $4.5 billion and $4.6 billion, and that earnings per share would fall in the range of 29 cents to 31 cents. The stock plummeted 12% following that release.
But Super Micro on Tuesday gave investors their first glimpse into fourth-quarter results, and those are also below expectations. Super Micro called for 40 cents to 50 cents in adjusted earnings per share on $5.6 billion to $6.4 billion in revenue. Analysts polled by LSEG had been looking for 69 cents in adjusted earnings per share on $6.82 billion in revenue.
The macroeconomic environment is likely to weigh on performance, the company said, following President Donald Trump’s announcement in early April of sweeping new tariffs on imported goods. CEO Charles Liang also said that some customers delayed purchases of data center technology in the latest quarter.
“We do expect many of those commitments to land in the June and September quarters, reinforcing my confidence in our ability to meet our long-term targets,” Liang said in the release. He added that “economic uncertainty and tariff impacts may have a short-term impact.”
Super Micro’s revenue grew 19% year over year during the quarter, which ended on March 31. Net income of 17 cents per share were down from 66 cents in the same quarter a year ago.
It’s been a treacherous past year for Super Micro. Prior to that, the stock had been on a tear due to the company’s position in the artificial intelligence market, selling servers packed with Nvidia’s graphics processing units.
Over the summer, short seller Hindenburg Research issued a report on the Super Micro, claiming it had found proof of “accounting manipulation.” In October, Ernst & Young resigned as the company’s auditor after raising concerns about internal control over financial reporting and other matters.
An independent special committee investigated but “did not raise any substantial concerns about the integrity of Super Micro’s senior management or Audit Committee, or their commitment to ensuring that the Company’s financial statements are materially accurate,” according to a statement.
In February, Super Micro filed an annual report for its 2024 fiscal year, which ended on June 30, helping to keep the stock from being delisted on Nasdaq. Staff from the exchange had informed Super Micro that the company was back in compliance with filing requirements, according to a statement.
As of Tuesday’s closing bell, Super Micro had gained 9% so far in 2025, while the S&P 500 index had declined by 4%.
Executives will discuss the results on a conference call starting at 5 p.m. ET.
This is breaking news. Please check back for updates.
Lisa Su, CEO of AMD, attends the Artificial Intelligence Action Summit at the Grand Palais in Paris on Feb. 10, 2025.
Benoit Tessier | Reuters
Advanced Micro Devices reported first fiscal-quarter earnings on Tuesday that topped expectations, and provided a strong forecast for current-quarter revenue.
Shares of AMD rose more than 4% in extended trading.
Here’s how the chipmaker did versus LSEG expectations for the quarter that ended March 29:
- Earnings per share: 96 cents adjusted vs. 94 cents expected
- Revenue: $7.44 billion vs. $7.13 billion expected
For the current quarter, AMD expects about $7.4 billion in sales with a gross margin of 43%, versus Wall Street estimates for earnings of 86 cents adjusted on $7.25 billion in sales.
AMD’s forecast also included $800 million in costs that the company said it would incur because the U.S. limited the export of some of the company’s artificial intelligence chips during the quarter.
The company reported net income of $709 million, or 44 cents per diluted share, versus net income of $123 million, or 7 cents per share, during the year-earlier period. Revenue grew 36% on an annual basis.
AMD is the second-place server central processing unit vendor, behind Intel, but its Epyc line of processors has been taking market share in recent years.
The company is also the closest competitor to Nvidia for “big GPUs,” or graphics processing units. Those are the kind of chips that are deployed in data centers by the thousands for building generative AI. It did $5 billion in AI GPU sales in the company’s fiscal 2024.
Both are reported in the company’s data center segment, which came in at $3.7 billion in sales, topping a StreetAccount estimate. Data center sales were up 57% on an annual basis, which the company attributed to demand for both Epyc processors and Instinct GPUs.
The company’s other major segment, Client and Gaming, includes chips for consumer devices such as laptops, gaming PCs and game consoles. The overall segment rose 28% on an annual basis to $2.9 billion. AMD said sales for its laptop and PC chips, which it calls client revenue, surged 68% year over year because of strong demand for chips called Zen 5 that the company released last summer.
Gaming sales, however, declined 30% on an annual basis, which the company attributed to a decrease in console chip revenue.
AMD’s embedded segment, which is mostly sales from the company’s 2022 acquisition of Xilinx, declined 3% on an annual basis to $823 million.
-
Sports3 years ago‘Storybook stuff’: Inside the night Bryce Harper sent the Phillies to the World Series
-
Sports1 year agoStory injured on diving stop, exits Red Sox game
-
Sports2 years agoGame 1 of WS least-watched in recorded history
-
Sports2 years agoMLB Rank 2023: Ranking baseball’s top 100 players
-
Sports4 years ago
Team Europe easily wins 4th straight Laver Cup
-
Environment2 years agoJapan and South Korea have a lot at stake in a free and open South China Sea
-
Environment2 years agoGame-changing Lectric XPedition launched as affordable electric cargo bike
-
Business3 years agoBank of England’s extraordinary response to government policy is almost unthinkable | Ed Conway