TikTok’s U.S. entity “is looking to secure multiple gigawatts of computing capacity,” CNBC reported on 23 October 2026, citing two people familiar with the discussions. “TikTok U.S. is looking for four gigawatts of compute power by the early 2030s, one source said.”
For scale, Meta describes its Hyperion site in Louisiana as scaling toward 5 gigawatts, according to our AI data-centre map. A gigawatt is a billion watts, a measure of electrical power that data centres use to size their capacity.
Business Pill · TOKENS PER WATT
A short Business Engineer explainer on tokens per watt: what an AI data centre actually produces. It teaches the general idea only and says nothing about TikTok or any company in this piece.
The key insight: As we read it, the new number is who is asking: per CNBC, TikTok’s U.S. entity, a social media company, is looking for four gigawatts of compute power by the early 2030s, one source said, close to the toward-5-gigawatt figure Meta gives for its Hyperion site.
The search. The company, officially TikTok USDS Joint Venture, “has sounded out opportunities to obtain capacity over the coming years,” according to two people familiar with the discussions who asked to remain anonymous, CNBC reported.
The size. One source put the target at four gigawatts of compute power by the early 2030s.
The cloud. A TikTok U.S. spokesperson told CNBC that the company “runs primarily on Oracle Cloud Infrastructure in the U.S. and Oracle will continue to be our primary cloud infrastructure provider.” CNBC notes that Oracle Cloud Infrastructure typically runs in facilities Oracle leases, and says it approached Oracle for comment.
The AI work. A TikTok job advert cited by CNBC says the company is “shipping a flagship B2B AI product at the frontier of large-model inference and creative automation.”
The algorithm. In January, TikTok U.S. said it would “retrain, test, and update” its recommendation algorithm on U.S. user data, with the algorithm “secured in Oracle’s U.S. cloud environment,” per CNBC.
Stated capacity, in gigawatts, as each source gives it: TikTok U.S. target of 4 GW of compute power by the early 2030s (one source, CNBC, 23 Oct 2026); Meta’s Hyperion scaling toward 5 GW (per Meta, via our AI data-centre map); CoreWeave’s 4.2 GW of locked-in power capacity (Goldman Sachs note, via our CoreWeave piece); 1.4 GW of total IT capacity in planning documents for the Dalby site where Anthropic signed the first of four buildings (Capital Brief). Different measures and dates; not summed.Stated gigawatts, as each source gives them: Meta’s Hyperion toward 5 GW (per Meta, via our AI data-centre map); CoreWeave’s 4.2 GW of locked-in power capacity (Goldman Sachs note, via our CoreWeave piece); TikTok U.S.’s target of 4 GW of compute power by the early 2030s (one source, CNBC, 23 Oct 2026); 1.4 GW of IT capacity in planning documents for the Dalby site (Capital Brief). Different measures; not summed.
What Was Posted on X
This might be the most underrated AI story of today.
According to CNBC, TikTok’s US business is looking to secure multiple gigawatts of computing capacity, with one source suggesting the company is targeting 4 GW by the early 2030s. I don’t think people understand how significant this is. We’re talking about a social media platform potentially looking to secure enough power and computing infrastructure to rival some of the largest AI data center developments in the country. For perspective, Meta’s massive Hyperion data center campus in Louisiana is targeting around 5 GW of capacity.
TikTok already relies primarily on Oracle Cloud Infrastructure, and the company has confirmed that $ORCL will remain its primary cloud provider in the United States. On top of that, TikTok is retraining its recommendation algorithm using American user data and is developing AI products focused on large model inference and creative automation. In other words, this isn’t necessarily just about supporting people scrolling through videos anymore. TikTok could be preparing for significantly more AI workloads which would require enormous amounts of computing power.
Now connect this to everything else we’ve been discussing. Microsoft, Amazon, Google, Meta, OpenAI, and xAI are already competing to secure power, GPUs, land, and data center capacity. Now we’re seeing another major technology platform potentially entering that same competition at the gigawatt scale. And remember, these companies cannot simply decide they want another gigawatt of computing capacity and have it available tomorrow. They need electricity, substations, cooling infrastructure, networking equipment, GPUs, memory, and facilities capable of supporting all of it.
This is exactly why I remain so bullish on the companies supplying the infrastructure behind AI. Whether we’re talking about $NVDA, $MU, $ORCL, $BE, or the broader networking and power ecosystem, the number of potential customers competing for limited resources continues to grow. What’s particularly interesting is that this demand is increasingly coming from companies outside the handful of frontier AI labs everyone focuses on. I think we’re still underestimating just how much infrastructure the AI economy is going to need over the next decade.
@MelvinInvests said this “might be the most underrated AI story of today.” The poster compares the reported four-gigawatt target, which a source only suggests, with Meta’s Hyperion campus, which they put at around 5 gigawatts, and argues the demand is increasingly coming from companies outside the frontier AI labs. The comparison and the argument are the poster’s; they read the same CNBC figures as a sign that new buyers may be entering the race for power and data-centre capacity.
Who Owns the Buyer
TikTok USDS Joint Venture was set up in 2026 under President Donald Trump’s executive order requiring TikTok’s U.S. business to be divested from ByteDance, CNBC reported. Its three managing investors, Silver Lake, Oracle and MGX, each hold 15%; the Dell Family Office is also an investor, and ByteDance kept 19.9%.
That makes Oracle both a 15% managing investor in TikTok U.S. and the cloud provider the company says will remain its primary one. CNBC did not report whether Oracle would supply the extra gigawatts.
Each figure as its source states it. They measure different things (a compute target, planned site capacity, contracted power, IT capacity) at different dates, so they are shown side by side and not added up.
Connecting the Coverage
Our AWS piece covered Reuters’ report of Amazon chief executive Andy Jassy saying “2027 is largely sold out” for Amazon Web Services’ computing capacity, the same week CNBC reported that TikTok U.S. has sounded out opportunities to obtain capacity.
Our SK Hyper piece reported Bloomberg’s account of a plan to build 15 gigawatts of data centre capacity in South Korea by 2035, with the first 5-gigawatt phase alone requiring $350 billion. Our CoreWeave funding piece covered a Goldman Sachs note that put CoreWeave’s locked-in power capacity at 4.2 gigawatts.
Our Oracle piece reported that Oracle said on 23 October 2026 it is committing $1 billion to its partner ecosystem over the next four years. Our AI data-centre map counts about 59.0 GW across the named projects in our coverage that carry a stated capacity, by our arithmetic, as an upper bound because it mixes operating, contracted and planned capacity.
What’s new: a social media company, not an AI lab or a cloud provider, is reported to be seeking compute at the scale of the largest campuses in our map.
The Structural Read
As we read it, the buyer and the supplier overlap: CNBC reports that Oracle is one of TikTok U.S.’s three managing investors with 15%, and a TikTok U.S. spokesperson told CNBC that Oracle will continue to be the company’s primary cloud infrastructure provider.
As we read it, the workload is changing: CNBC cites a TikTok job advert for a “flagship B2B AI product at the frontier of large-model inference and creative automation,” on top of retraining the recommendation algorithm on U.S. user data.
As we read it, the request lands in a tight market: Reuters reported Andy Jassy saying “2027 is largely sold out” for Amazon Web Services’ computing capacity, as covered in our AWS piece.
A TikTok U.S. spokesperson, to CNBC, 23 October 2026
“runs primarily on Oracle Cloud Infrastructure in the U.S. and Oracle will continue to be our primary cloud infrastructure provider.”
Three Implications
FOR ORACLE CNBC reports that Oracle holds 15% of TikTok U.S. as a managing investor and that the company names Oracle as its primary cloud; CNBC said it had approached Oracle for comment.
FOR THE COMPUTE MARKET CNBC frames the search as part of “a global scramble for digital infrastructure deals,” with companies across the world looking to secure more compute capacity as AI workloads grow.
FOR READERS OF GIGAWATT FIGURES The four-gigawatt target is one source’s figure for the early 2030s; it is a target, not capacity that has been contracted or built.
The framework opens: “Big Tech’s hyperscalers are racing to build gigawatt-scale AI data centers, with five facilities set to exceed 1GW by 2026, each requiring a nuclear reactor’s worth of power.”
As we read it, CNBC’s report adds a non-hyperscaler to that race: a social media company sizing its compute needs in gigawatts. The framework was published in December 2025 and does not discuss TikTok.
What Is Not Established
The four-gigawatt figure comes from one source, and CNBC describes the company as looking for capacity, not as having signed for it. CNBC did not report a cost, a site, a supplier for the added capacity or a timeline beyond “the early 2030s.” Oracle had not commented in the article we read.
Business Engineer Framework
The State of AI Data Centers
A Business Engineer map of who is building gigawatt-scale AI data centres and where power becomes the constraint.
CNBC reports that TikTok’s U.S. entity is looking to secure multiple gigawatts of computing capacity, with one source putting the target at four gigawatts by the early 2030s. The company says Oracle, one of its three 15% managing investors, will remain its primary cloud provider.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. This piece rests on CNBC’s report of 23 October 2026, which cites two people familiar with the discussions and a TikTok U.S. spokesperson, and on our own earlier pieces for the comparison figures. We read CNBC’s article in full. We did not contact TikTok U.S. or Oracle; CNBC said it had approached Oracle for comment.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
