GW

Where the record uses it

• Campus Hyperion
• Campus Colossus 2
• Campus Stargate UAE
• Campus Stargate Michigan / The Barn
• Campus Project Matador / HyperGrid
• Scorecard indicator ERCOT raw queue / live load

Sources that use it

• Dan Wang, Breakneck: China’s Quest to Engineer the Future (2025), read in full (the “engineering state” vs “lawyerly society” frame; the US offshore-wind ledger Wang dates to 2024, which its source, the Department of Energy’s Offshore Wind Market Report 2023 Edition, dates to 31 May 2023 — 42 MW operating, 932 under construction, 20,978 in permitting review; Cape Wind’s sixteen years of litigation; the only US nuclear plant built this century, “fifteen years and $30 billion”, the expected cost of the eleven reactors China’s nuclear authority approved in August 2024, ch. 2, p. 30, his note citing Bloomberg of 20 August 2024; the nuclear-powered Meta data center halted by a rare species of bee near the site, ch. 6, p. 203, his note citing the Financial Times of 4 November 2024; China’s 76 GW of wind installed in 2023 to America’s six; “the United States brought lawyers to a technology fight” and “rather than realizing its own Sputnik moment, the United States triggered one in China”; Nicholas Bagley’s “The Procedure Fetish” as Wang quotes it, Bagley’s word “available” dropped from “ventilate every available option” and his hypothetical extreme of a spectrum given as what the federal process requires; the book quotes Bagley’s own text from the article’s row; technology as tools, blueprints and process knowledge, the last existing “mostly in people’s heads” and embodied by an experienced workforce, ch. 3, with the NNSA’s $69 million relearning of Fogbank as the American cautionary tale, p. 74; the munitions asymmetry — nearly 1,800 commercial ships under construction in China to five in the United States in 2022, ch. 6, p. 204, his note citing the Congressional Research Service’s report on U.S. commercial shipbuilding of 15 November 2023, and UNIDO’s 45 percent of world industrial capacity by 2030, pp. 204–206, a forecast his notes do not source — and “algorithms alone will never win a battle”; “The United States, Europe, Japan, South Korea, Taiwan, and all other high-income states combined add up to 38 percent of capacity”, the sentence after the UNIDO forecast; “In two days, Ukraine could fire as many shells as the United States makes in a month”, p. 204, his note giving about 7,000 Ukrainian shells a day against 14,400 American shells a month from the New York Times of 29 May 2024 and Bloomberg of 7 June 2024; the Introduction: “For six years, I lived through a period of economic dynamism that gave way to smothering political repressiveness … I tracked the expanding web of US restrictions on Chinese tech companies, as well as their struggle to escape from American restraints”; the question closing the paragraph before the shipbuilding figures, ch. 6: “Which would you rather have: software or hardware?”; the full text read carries no page breaks, and the pages given are those its endnotes print for each noted passage; on Cape Wind, ch. 2: “the lawyerly society is great at protecting the wealthy”; on the controls: “scorched China’s most dynamic companies without killing them” (ch. 6, Fortress China); ch. 2: by 2025 China had caught up with the United States in the number of nuclear plants, fifty-five to fifty-four, with thirty-one under construction against America’s one; ch. 6: “American companies are not so much constrained on computing power as they are on electrical power” and “what China lacks in technological sophistication, it might make up for in electrical power”; ch. 6: “They ignored Beijing’s beseeching to buy from domestic vendors for the simple reason that Chinese technologies were not good enough”; ch. 6: “In 2024, the United Nations Industrial Development Organization forecast that China will have 45 percent of the world’s industrial capacity by 2030”). In the Sources
• Dwarkesh Patel with Gavin Leech, The Scaling Era: An Oral History of AI, 2019–2025 (Stripe Press, 2025) (curated interviews from the Dwarkesh Podcast, 2019–2025, the book’s own note on its text being “The interview excerpts have been edited for clarity and flow”, so each quotation here is the editors’ text, and where the recording differs the chapter cites the podcast’s own row; Amodei’s unpublished 2017 OpenAI memo “The Big Blob of Compute”; Zuckerberg, 2024, as the editors print him: “I think it’s worth investing $100 billion plus in building the infrastructure”, where the podcast’s transcript and the recording have the $10Bs or $100B+ and “and assuming that if it keeps going, you’re going to get some really amazing things… I don’t think anyone in the industry can really tell you for sure that it will continue scaling at that rate” (chapter 5, the ellipsis for “that make amazing products”); Zuckerberg, 2024, energy the coming bottleneck: “No one has built a 1 GW data center yet”, the recording’s a single gigawatt, and of new power plants and transmission lines, “You’re talking many years of lead time”; Dylan Patel, 2024, in the editors’ text and found on no published recording or transcript of the podcast: “In 1990, ExxonMobil was not laying fiber. In 2025, Microsoft is building data centers”; “prior bubbles didn’t have the most profitable companies that humanity has ever created investing in them. They were debt financed instead. AI is not debt financed yet”, the editors’ wording of the recording’s they were debt financed. This is not debt financed yet; Dylan Patel, 2024, in the editors’ text and found on no published recording or transcript of the podcast: “If human capital is infinite, which is what AGI is, then theoretically the returns of AGI are infinite. If I’m Mark Zuckerberg or Satya Nadella, I now have potentially infinite returns — if I get there first” (chapter 6, the book’s closed dash spaced); Aschenbrenner, 2024: a 10 GW cluster “a couple hundred billion”, the recording’s maybe before it dropped by the editors, the payoff test “you have to get to, like, $100 billion a year in revenue”, the podcast transcript’s form of the recording’s eventually you got to get to like 100 billion a year; Aschenbrenner, 2024: “Let’s estimate the GPT-4 cluster at around $500 million”, and of the $100 million figure, “That’s just the rental price. If you’re building the biggest cluster, you have to build and pay for the whole cluster”, the second sentence the podcast transcript’s form of the recording’s you got to build the whole cluster, you got to pay for the whole cluster; Aschenbrenner, 2024, on buying an aluminum smelter’s gigawatt: “a lot of these power contracts are locked in long term. Also, people don’t like it”, the recording’s people don’t like things like this; Aschenbrenner, 2024: “Would you do the Manhattan Project in the UAE?”, “Once the cluster is there, it’s much easier for them to exfiltrate the weights”; Shulman: a percentage point of search share “a billion-dollar matter”, $50–70 trillion of annual wages as the prize, and “the chance of advanced AI happening is relatively concentrated in the next 10 years [between 2024 and 2034] compared to the rest of the century, because we can’t keep going with our current rapid redirection of resources into AI. That’s a one-time thing”; Amodei: datacenter networks “cost as much as aircraft carriers”, a forecast on the recording (we’re anyway heading to a world where), which the editors set as already; Legg’s 50 percent by 2028; Sholto Douglas, March 2024, as the editors print him: “That is exactly the problem. But the key issue is that your base task solve rate is 90 percent. If it was 99 percent, it wouldn’t be a problem”; of the memo, “an unpublished memo he wrote at OpenAI in 2017—before scaling laws, before the Transformer, and before LLMs”, the editors’ order, where the recording they draw on has Amodei saying of the transformer paper It might have been just before. It might have been just after; chapter 5, “Inputs”, the same exchange as the book edits it: “You can put the clusters in the US and in allied democracies. Once you put them in authoritarian dictatorships, you create this irreversible security risk”; “Another thing: They can just seize the compute”; “Suppose we put 25 percent of the compute capacity in these Middle Eastern dictatorships. Say they seize that. Now the US has a 3-to-1 compute ratio.”). In the Sources
• Adani Group, “Google Breaks Ground on India AI Hub,” 28 April 2026 (nearly 1 GW, Visakhapatnam; ~$15 billion over 2026–30). In the Sources
• AEP Ohio, “New process advances 5.6 GW of data center projects,” 13 February 2026 (the page’s own title is “AEP Ohio Updates PUCO on Data Center Load: Figures Show Tariff is Working”; its figures as of 12 February 2026: of about 30,000 MW requested before the tariff, “13,022.7 MW asked and paid for a study,” 5,642 MW signed after the tariff took effect beside “12,219 MW of data center contracts signed before the tariff went into effect,” 17,861 MW in all “scheduled to come online progressively through 2035”) In the Sources
• AMD, “AMD and Anthropic Announce Strategic Partnership to Deploy Up to 2 Gigawatts of AMD Instinct MI450 Series GPUs,” 22 July 2026 (up to 2 GW for Anthropic, the first gigawatt beginning in H1 2027; AMD may invest up to $5 billion in Anthropic; “AMD has committed to make a strategic equity investment of up to $5 billion in Anthropic in the future”); OpenAI up to 6 GW and Meta up to 6 GW, each in its own release, at the second and third addresses. In the Sources

And 114 more: search the Sources.

Related terms

• Gigawatt