U.S. data center investment now exceeds all past booms, with about $10 trillion in projected spending and an economic impact of nearly 4% of GDP, almost twice that of 19th-century railroad construction (about 2% of GDP, $173 billion in today's dollars). Texas ($638 billion), Virginia ($249 billion), Ohio ($169 billion) and Georgia ($160 billion) are projected to receive the most spending over three years. New data center construction has more than doubled to $85 billion a year while other construction fell 8%, with data centers adding $48 billion since 2024 as new houses and apartments dropped $29 billion, and economists warn the narrow, fragile growth base and unproven AI profitability could lead to sharp cutbacks. Several states are rethinking tax breaks granted before the boom, as tax revenues and energy costs take a toll.
Summary written by Upside AI from the original article. Read the source for full details.
Read the original on orlandosentinel.com ↗
Keywords
data centersAI investmentartificial intelligenceUS
Open a keyword to see other news on the same topic.
Related tickers and likely impact
NVDA
▲ Likely positiveNvidia
Relevance: medium
| Market | US stocks |
|---|---|
| Sector | Information Technology |
| Industry | Semiconductors |
| Market cap | $5.2T |
| Close on 10/21/2026 | 221.87 |
| Change that session | +2.49% |
AMZN
▲ Likely positiveAmazon
Relevance: medium
| Market | US stocks |
|---|---|
| Sector | Consumer Discretionary |
| Industry | Broadline Retail |
| Market cap | $2.7T |
| Close on 10/21/2026 | 262.28 |
| Change that session | -0.13% |
GOOGL
▲ Likely positiveAlphabet Inc. (Class A)
Relevance: medium
| Market | US stocks |
|---|---|
| Sector | Communication Services |
| Industry | Interactive Media & Services |
| Market cap | $4.3T |
| Close on 10/21/2026 | 334.67 |
| Change that session | -1.65% |
META
▲ Likely positiveMeta Platforms
Relevance: medium
| Market | US stocks |
|---|---|
| Sector | Communication Services |
| Industry | Interactive Media & Services |
| Market cap | $1.9T |
| Close on 10/21/2026 | 698.43 |
| Change that session | +0.60% |
MSFT
▲ Likely positiveMicrosoft
Relevance: medium
| Market | US stocks |
|---|---|
| Sector | Information Technology |
| Industry | Systems Software |
| Market cap | $4.0T |
| Close on 10/21/2026 | 532.81 |
| Change that session | +0.18% |
Impact is scored by AI from the headline and may be wrong. It is not investment advice.
Related news
- Data centers are now the largest financial bet in US history
US data center investment, projected at $10 trillion, is now the largest financial bet in US history, surpassing the railroad expansion, and its economic impact is nearly 4% of GDP, almost twice that of railroads, according to a Columbia Business School report. The boom is crowding out other construction: data center construction value has more than doubled to $85 billion a year, while all other construction fell 8%, and new houses and apartments dropped $29 billion as data center value rose $48 billion since 2024. Many states that granted tax exemptions earlier are now rethinking their approach; Nevada trimmed abatements and will make data centers cover energy and water costs. Texas ($638 billion), Virginia ($249 billion), Ohio ($169 billion) and Georgia ($160 billion) are projected to see the most spending over three years. Officials and economists note AI profitability has yet to catch up with the spending, and warn that growth relying on this narrow base is fragile.
- Data centers are now the largest financial bet in US history
Data center investment is surpassing every previous US boom, with a Columbia Business School report projecting $10 trillion in spending and an economic impact of nearly 4% of GDP, almost twice that of the railroad expansion. New data center construction value has more than doubled to $85 billion annually, while all other construction value fell 8%, and new houses and apartments dropped $29 billion as data center value grew $48 billion since 2024. Texas, Virginia, Ohio and Georgia are projected to account for the largest costs, and Nevada has trimmed tax abatements and required data centers to cover energy and water demands. Analysts say AI profitability has yet to catch up with the buildout, and economist Josh Bivens warns that growth relies heavily on one fragile sector.
- What the Data Center Debate Gets Wrong
- Data Centre Stock Falls 40 % as Investors Lose Faith in AI Story
Shares of NextDC, one of Australia's largest data centre builders, fell 40% after the company said delays at its flagship Melbourne site would leave revenue forecasts short. The delays followed community concerns over its M3 expansion, which led Jemena to pause construction of 22-metre power poles in Footscray. The article points to wider strain in the sector: Firmus cancelled its Australian listing for lack of demand, Nscale is struggling to attract buyers for a Wall Street listing, and CoreWeave dropped 15% in a day. It also notes that investment could reach US$10 trillion (A$14.25 trillion) by 2032, while higher borrowing and construction costs and the need for AI firms to lift revenue 80% a year for six years add pressure.
- Nvidia turns the AI factory into a product , and agents into the workload
At GTC Berlin, Nvidia CEO Jensen Huang focused on deployment speed, power, security and enterprise data, describing AI as a manufacturing industry. Nvidia introduced DSX POD, a containerized reference design for complete Vera Rubin AI factories that it says can be deployed at a prepared site in as little as 14 days, versus six months, though the figure assumes the site is ready. Huang said a 1-gigawatt AI factory costs about $50 billion and earns roughly $40 billion a year, so a one-quarter delay loses about $10 billion in revenue. The DSX platform adds DSX Flex, which automates how data centers defer work for grid operators, and cooling that saves an estimated 54 million liters of water per day at 2 gigawatts. Nvidia also launched an OpenShell Reference Architecture for running agents safely and expanded its Agent Toolkit.
- AI's billion-dollar question: When will the spending pay off?
Speakers in this discussion argue AI infrastructure spending faces a timing gap, since months can pass between when computing hardware ships and when it starts earning revenue, leaving a large amount of capital tied up. As an example, a buyer of $10 million in Nvidia chips may need six months to get them running while that money sits on its balance sheet. They also point to a wider gap between total AI spending and the revenue it generates, and note that some reports suggest the size of returns may depend on technological innovation. Easing supply chain bottlenecks, they say, would help companies bring computing capacity online faster.
Related markets
Latest news
- OpenAI nears up to $30 billion funding round at $1.25 trillion valuation led by UAE investors
OpenAI is nearing a funding round of up to $30 billion at a $1.25 trillion valuation, led by UAE investors alongside Coatue, with that valuation excluding the new capital. The company aims to get signatures by the end of next week, though the final size is still being determined. The raise is positioned as an “opportunistic” bridge ahead of a potential IPO, which CFO Sarah Friar has told employees will likely come in 2027. OpenAI is also negotiating a commercial deal that would secure more computing capacity from the UAE. The move comes amid choppy tech markets, with jitters around CoreWeave and concerns about financing as interest rates rise, while Anthropic's expected IPO will test investor appetite. OpenAI has filed confidentially to go public.
- Here's what is spooking investors as Meta stock stumbles
RBC Capital Markets analyst Brad Erickson says Meta shares are under pressure after earnings, with the language in the company's filing unsettling investors. He explains that Meta is issuing non-voting stock to fund billions in AI spending while preserving CEO Mark Zuckerberg's voting control. Erickson said the dilution level cited, nearly 800 billion, is probably well above the actual investment, and that the CFO described that language as boilerplate. He also noted the move lets Zuckerberg keep control and stay in California while a tax situation is being worked out.
- A lesson from poker tells us not to fold on Meta after its earnings sell-off. Here’s why
Meta's strong quarterly results were overshadowed by concerns over AI spending, sending shares down 7% on Thursday. Bears point to dilution risk, as Meta plans new non-voting Class C shares to fund AI infrastructure and free cash flow turned negative in the quarter. A New Mexico judge also ordered Meta to pay a $25 billion penalty over the Cambridge Analytica scandal, which the company plans to appeal, and Meta booked a $9.89 billion legal settlement charge in the quarter. Bulls argue the core business remains strong, with the new Muse app off to a strong start, and CFO Susan Li said in-house content models are outperforming prior versions in offline tests.
- Resideo Announces Expected Q1 Software Update to Honeywell Home Smart Thermostats for Compatibility with Apple HomeView
Resideo plans to release a software update in the first quarter of 2027 that will let its existing Honeywell Home ElitePRO, X7S and X8S smart thermostats support the new Adaptive Temperature Control and Clean Energy Guidance features tied to Apple's HomeView offerings. Homeowners can get these features without replacing devices or paying for extra professional installation. The thermostats already work with Apple Home, Google Home, Amazon Alexa and other ecosystems and are Matter certified. Resideo says it works with more than 100,000 professionals worldwide, who bring its products into more than 15 million installations annually.
- Why Nokia Sank today
Nokia shares fell 15.4% on Thursday, as of mid-afternoon, in a sector-wide sell-off of AI-related stocks, even though the company beat earnings expectations. Third-quarter revenue rose 9.7% to 5.15 billion euros, adjusted EPS rose 50% to $0.09, and operating margin expanded 220 basis points. AI and cloud customer revenue jumped 123% and order intake for that segment rose 118%, while mobile networks grew only 5%. Nokia reiterated but did not raise its full-year 2026 guidance, and the article says high expectations after the stock's rally this year, along with fears about the sustainability of AI spending, weighed on shares.
- Google Is Paying SpaceX $11 Billion a Year for Compute. Here's the Concentration Risk Nobody Is Pricing In.
Google has started paying SpaceX $920 million per month, or $11 billion per year, to lease xAI's cloud infrastructure through 2029, rather than building more data centers to fill its shortage of Nvidia GPUs. According to Synergy Research Group, Google held 15% of the global cloud infrastructure market in the second quarter of 2026, placing third behind AWS (28%) and Microsoft Azure (20%), while Google Cloud revenue rose from $19.2 billion in 2021 to $58.7 billion in 2025. The deal gives Google access to a cluster of roughly 110,000 GPUs, and Alphabet owns a 4%–5% stake in SpaceX. The author warns the arrangement adds concentration risk: Google is effectively funding xAI's growth into a cloud competitor, it reinforces Nvidia's dominance, and the cluster sits at xAI's centralized facilities, including Colossus in Tennessee, where power failures or disruptions could affect many tenants.
Get the briefing on Telegram or Zalo
Join the Upside Telegram channel or Zalo group for the Vietnam stocks and gold briefing at 8:00 and 19:00 (Vietnam time) on trading days, plus a weekly wrap on Saturday morning.