Technology companies’ spending on artificial intelligence chips has plunged Meta, Alphabet and Tesla into an “arms race” | Real Time Headlines

Yuan CEO Mark Zuckerberg Already assembling a large inventory Nvidia wafers, spending billions of dollars so that his company can develop and train advanced artificial intelligence models.

But he also said the hype around artificial intelligence may have led to too much investment.

“I think a lot of companies right now are overbuilding, and when you look back, you think, oh, we probably all spent billions of dollars more than we should have spent,” Zuckerberg said. podcast Meet Bloomberg’s Emily Chang this week.

He is not the only one to express this view.

exist alphabetical During Wednesday’s earnings call, C.E.O. Sundar Pichai said his company was likely investing too much in artificial intelligence infrastructure, which primarily consists of Nvidia’s graphics processing units (GPUs). But he sees no alternative.

“When we go through a curve like this, the risk of underinvesting is much greater than the risk of overinvesting,” Pichai said.

In addition to Meta and Alphabet, Nvidia is also expanding its business through the following channels: Microsoft, Amazon, Oracle and Tesla, all of whom have publicly declared that investing in artificial intelligence is a core priority for this year and the foreseeable future. NVIDIA income It has more than doubled for three consecutive quarters and is expected to more than double in the current period.

Alphabet and Tesla highlighted the cost of building out their artificial intelligence in earnings calls this week, and investors expect to hear more news next week when Microsoft, Amazon and Meta report results.

Yuan Debut On Tuesday, its latest Llama AI model. The model is called Llama 3.1 and comes in three different versions, one of which is Meta’s largest and most powerful AI model to date. Meta’s insistence on open source means that outside developers can access the technology for free, even though the company has invested heavily in the underlying infrastructure.

Zuckerberg said in the podcast with Zhang that the company is “making rational decisions” in investing in artificial intelligence despite the high cost.

“Because the disadvantage of falling behind is that you won’t be able to master the most important technologies in the next 10 to 15 years,” Zuckerberg said.

In Pichai’s view, even if Alphabet invests too much, the infrastructure “is very useful to us.”

Nvidia shares are up 131% this year after surging 239% in 2023. briefly transcend Their market value in June.

More than 40% of Nvidia’s revenue comes from Microsoft, Amazon, Google and Oracle, and these companies’ public cloud products require a large number of GPUs. Although these companies are some of the best capitalized in the world, investors are worried about the massive inventory.

David Cahn, a partner at venture capital firm Sequoia Capital, wrote in a blog post last week that the spending is being driven by competitiveness and Follow game theory dynamicscreating a “cycle of competitive escalation.”

“The cloud giants see artificial intelligence as both a threat and an opportunity, and they don’t have enough time to wait and see how the technology develops,” Kahn said. wrote. “They must act now.”

Kahn calculated that the tech industry would need $600 billion in annual AI revenue to achieve this goal. prove legal All the money is spent on data centers and chips.

On Wednesday, Kahn went on to say that comments from Zuckerberg and Pichai about limiting downside risks supported his theory.

“The CEOs of Google and Meta have both agreed with my AI arms race narrative in the past 24 hours: AI capex is driven by game theory and FOMO versus actual revenue/usage,” Kahn release It’s LinkedIn.

Nvidia said demand for its latest generation of artificial intelligence chips, Blackwell, which will begin shipping later this year, will remain strong. But it begins to address investors’ questions about investment returns as growth inevitably slows due to historically difficult comparisons.

Nvidia chief financial officer Colette Kress told investors in May that the company Already calculated When a cloud provider spends $1 to buy an Nvidia-based server, it can be rented out for $5. Goldman Sachs analysts said in a recent report that Nvidia is looking to share such data points to boost investor confidence.

Tesla CEO Musk Said in his company Earnings Conference Call On Tuesday, “Demand for Nvidia hardware is so high that it is often difficult to obtain GPUs.” Tesla said it spent $600 million on artificial intelligence capital expenditures in the quarter, mainly investing in autonomous driving and humanoid robots.

Musk said Tesla is focusing on developing its own Dojo supercomputer because Nvidia chips are expensive and difficult to obtain.

“I think we have no choice because the demand for Nvidia is so high,” Musk said. “Obviously they have an obligation to raise the price of GPUs to whatever level the market will bear, which is very high.

watch: Investors should look forward to earnings catalysts for Tesla