Proposed artificial intelligence capacity additions continue to mount, and Nvidia (NASDAQ:NVDA) appears poised to benefit the most from the trillions of dollars being poured into these initiatives over the next few years, according to Barclays.
Barclays’ methodology finds that each gigawatt of AI capacity requires $50B to $60B in total cost to build. They also calculate that a single gigawatt of capacity requires about 500,000 graphics processing units.
Earlier this week, Nvidia announced a plan to invest up to $100B in Microsoft (MSFT)-backed OpenAI to build and deploy at least 10 gigawatts of AI data centers with Nvidia systems. Using Barclay’s formula, this one project could cost up to $600B and require 4M to 5M GPUs.
“With the wave of announcements that have come over the last 6-9 months, we now estimate over $2T of planned spend at ~40 GW of power in total,” said Barclays analyst Tom O’Malley in an extensive investor note on Thursday. “Within that, we attribute ~65-70% to compute & networking with more deals likely in the pipeline, which starts to make the updated guidance of $3-4T look much more real.”
“We see this as a positive for all accelerator names, Broadcom (AVGO) and AMD (AMD), but we see this largely flowing into the NVDA P&L over the next 5+ years, moving numbers materially higher and making this the most attractive name in our space,” he added.
Barclays reiterated its Overweight rating on Nvidia and increased its price target by 20% to $240 from $200.
Of the 40 GW of AI capacity slated to come online globally over the next four to five years, some of the most substantial U.S. projects include Meta Platforms (META), 2 GW; Stargate, 10 GW with Oracle (ORCL), SoftBank (OTCPK:SFTBY)(OTCPK:SFTBF) and OpenAI; and multiple projects by CoreWeave (CRWV), Microsoft (MSFT), Amazon (AMZN) and Google (GOOG)(GOOGL) with capacities less than 1 GW. There are also several large-scale capacity additions planned for Saudi Arabia, South Korea, France, Abu Dhabi, Taiwan, Australia and the U.K.
In addition to hyperscalers and semiconductor firms benefitting from this buildout, several U.S. power and utility companies are poised to take advantage as well, according to Barclays. These include Entergy (ENT), WEC Energy Group (WEC), The Southern Company (SO), Evergy (EVRG), Ameren (AEE), NiSource (NI) and Alliant Energy (LNT). With these additions likely leading to higher energy prices, Barclays believes NRG Energy (NRG) and Talen Energy (TLN) are also preferred names.