Skip to main content
The Markets by Proactive
Go to Proactive UK
Proactive UK has moved. Proactive’s coverage of London’s small caps continues on proactiveinvestors.com Go there →
Advertisement
The Markets
by Proactive
Proactive UK has moved.
Coverage of London’s small caps continues on proactiveinvestors.com
Go to Proactive UK
The Markets
by Proactive
Proactive UK has moved.
Small-cap coverage continues on .com
Go to Proactive UK
Advertisement
The Markets
by Proactive
Proactive UK has moved.
Small-cap coverage continues on .com
Go to Proactive UK

Energy

Nuclear energy stocks slide on DeepSeek AI read-across

Shares in nuclear energy stocks that have enjoyed a boost from their new popularity with data centre owner were selling off after the emergence of a less power-hungry artificial intelligence model from Chinese startup DeepSeek.

This led to a widespread sell-off among chipmakers and other semiconductor companies in Europe, the US and Asia on Monday.

US nuclear companies were also hit, with new-generation reactor developers Oklo Inc (NYSE:OKLO) and Nuscale Power Corp (NYSE:SMR) dropping 16.3% and 13.7% in pre-market trading.

Shares in the pair had surged 271% and 800%-plus over the past 12 months, as Google, Amazon, Microsoft and Meta announced investments in nuclear developments to power their AI data centres.

Constellation Energy Corp (NASDAQ:CEG) was down 11.4% and Vistra Corp (NYSE:VST) was 11.2% lower, having respectively climbed 187% and 371% since a year ago.

DeepSeek launched a free open-source AI assistant R1 last Friday, with its large language model (LLM) matching those of OpenAI but at a tenth of the cost and requiring a fraction of the data, and therefore energy.

A research paper from the Hangzhou-based startup calculated the training costs of one of its models at less than $6 million.

Analysts at Jefferies said the market was "naturally" worrying about demand growth in computing power.

They said they believe success for DeepSeek (DS) could drive two possible industry strategies, the first being to continue pursuing more computing power to drive even faster model improvements, and the second to refocus on efficiency and return on investment, meaning lower demand for computing power from 2026.

"With a generous capital market, overseas AI companies have been going after model improvement at all costs. But DS could prompt investors to ask hard questions about these computing power investments. Therefore, US AI players' mgt could be under more pressure to justify further raising AI capex in 2026."

The analysts said the AI supply chain, ie those who make GPU, server ODMs, PCB and liquid cooling, is "highly vulnerable to a de-rating", with producers of ASIC, HBM, power DCs "more resilient".

Advertisement
The Markets
by Proactive
Proactive UK has moved.
Small-cap coverage continues on .com
Go to Proactive UK