Chinese AI startup Deepseek stunned the world with the release of its R1 model, which appears to perform nearly as well as leading models from Google and Openai, despite the company’s claim that it used a relatively modest number of GPUs to train it. .
Deepseek’s relative efficiency has pundits and investors questioning whether AI really needs the massive hardware spending that everyone had predicted. And that can change data center demand—and the energy needed to power them
The company claims it ran 2,048 NVIDIA H800 GPUs for two months to train a slightly older model, a part of the compute that Openai is rumored to use.
Few companies are as exposed as NVIDIA, whose share price was down 16% at press time. Perhaps even more tangible are the startups and power producers that are betting big on new nuclear and natural gas capacity.
Nuclear power, in particular, has been on the verge of a renaissance for years, fueled by advances in fuels and reactor designs that promise to make a new generation of power plants safer and cheaper to build and operate. . Until now, there was little reason to flash forward. Nuclear is still expensive compared to gas, solar and natural gas. Plus, the next-generation core has yet to be tested on a commercial scale.
The increased demand for energy from AI changed the equation. With data centers projected to consume as much as 12% of all US electricity—more than triple their share in 2023—and predictions of AI-powered data centers by 2027, technology companies are racing to secure new supplies, and throwing billions of dollars at the problem. Google has committed to buying 500 megawatts of capacity from nuclear startup Kairos, Amazon led a $500 million investment in another nuclear startup, X-Energy, and Microsoft is working with Constellation Energy on a $1.6 billion renovation of a reactor in three mile island.
What if the problem is overwhelming?
There is no hard and fast rule that suggests the only way to improve AI performance is to use more accounts. For a while, that tactic worked well, but lately, more calculation hasn’t yielded the same results. AI researchers have been scrambling for solutions, and it’s possible that Deepseek found one for its R1 model.
Not everyone is convinced, of course.
“While Deepseek’s achievement may have been staggering, we question the notion that its feats were accomplished without the use of advanced GPUs,” wrote Citigroup analyst Atif Malik.
However, the story suggests that even if Deepseek is hiding something, someone else will surely find a way to make him cheaper and more efficient. After all, it is easier and potentially faster to force a few PhDs to develop better models than it is to build new power plants.
The current wave of new reactors isn’t scheduled to come online until 2030, and new natural gas plants won’t be available until the end of the decade at the earliest. In that context, tech companies’ energy investments appear to be hedges in case their software bets don’t pan out.
If they do, expect tech companies to scale up their energy ambitions. When given the choice between spending billions on physical assets or software, tech companies almost always chose the latter.
Where will that leave nuclear startups and energy companies? That depends. Some may be able to produce power at a low enough cost that it won’t matter if AI power needs EBB. The world is electrifying, and even before the AI bubble started to burst, the demand for electricity was expected to increase.
But absent demand from AI, those cost pressures will probably increase. Wind, solar and batteries are cheap and getting cheaper, and they are inherently modular and mass-produced. Developers can roll out new renewable plants in phases, providing electricity (and revenue) before the entire project is complete while providing a control over their future in the face of uncertain demand. The same cannot be said for a nuclear reactor or a gas turbine. Tech companies know this, which is why they’ve been quietly investing in renewables to power their data centers.
Few people predicted the current AI boom, and it’s unlikely anyone knows how the next five years will play out. As a result, the safest bets in energy will probably flow to proven technologies that can be rapidly deployed and scaled according to a rapidly evolving market. Today, renewables fit that bill.
