By Zaeem Shoaib
As Big Tech ramps up spending on artificial intelligence, companies are increasingly looking overseas, investing billions of dollars to build out AI infrastructure.
So far this year, Microsoft and Amazon have earmarked more than $40 billion combined for investments in AI-related and data center projects worldwide.
Broadly, big tech companies are looking to "spread their wings" to international markets, Wedbush analyst Daniel Ives told The Wall Street Journal. "This is an AI arms race as Microsoft, Amazon and others skate to where the puck is going with this tidal wave of spending on the doorstep."
DA Davidson analyst Gil Luria expects these companies to spend more than $100 billion this year on AI infrastructure. Spending will continue to increase in line with demand, Luria said.
Ives expects significant continued investment in AI infrastructure by tech companies over the next 10 years, "This is a $1 trillion spending jump ball over the next decade."
Microsoft has plans for more than $16 billion in investments over the next several years spread between France, Germany, Japan, Malaysia, Spain and Indonesia.
Amazon, meanwhile, has planned infrastructure investments of $15 billion in Japan, $9 billion in Singapore, $5 billion in Mexico, and $1.3 billion in France.
"As the rest of the world migrates more to the cloud, there's more need to have these data centers built in the region where that shift is happening," Luria said.
The overall growing demand for generative AI, as well as the desire from new ventures and startups to incorporate this functionality into their businesses, has been pushing tech companies to expand their capabilities quickly. Companies "need those data centers full of Nvidia GPUs to be local" due to faster processing times as well as data privacy and security, Luria said.
Meanwhile, Google parent company Alphabet has kept most of its AI and data center-related investments confined to the U.S. Wedbush's Ives expects Alphabet, as well as Meta Platforms, to aggressively expand overseas, following the lead of Microsoft and Amazon, who are known as leaders in cloud computing. Luria said for these companies, expanding their data center capacity is "part of the natural progression of their business."
But Luria said Meta tends to put data centers where it can expeditiously deliver their services tied to its social media platforms, so it's less clear that they need to build an overseas data-center infrastructure similar to Microsoft and Amazon.
"As long as my Instagram feed loads quickly, I don't really care where the data center is," Luria added. Still, Meta expects to invest $35 billion to $40 billion this year to support AI capabilities.
Josh Beck, Managing Director of Equity Research at Raymond James, believes that as Meta continues to build out its large language model LLAMA, it "opens the door on commercialization." He thinks Meta could potentially bundle its AI services with compute capabilities to expand its services but doesn't see it directly competing for large enterprise customers.
Write to Zaeem Shoaib at zaeem.shoaib@wsj.com
(END) Dow Jones Newswires
May 23, 2024 13:56 ET (17:56 GMT)
Copyright (c) 2024 Dow Jones & Company, Inc.
Comments