The research firm founded by Dylan Patel, whose estimates of chips, datacenters and rents run through the chapters. It sells its research into the buildout it counts. Chapter 7.
From The AI Compute Race. Read the book
Where the book uses it
- Chapter 7 — The Double Blockade
Two governments deciding one trade: Washington’s export rules and Beijing’s instructions at the border, and what each did to the chip business it meant to shape.
Where the record uses it
- Dated test Chapter 7, the ecosystem beneath it, due 31 December 2027
- Person Patel, Dylan
Sources that use it
- Anthropic, “Anthropic raises $30 billion Series G funding at $380 billion post-money valuation,” 12 February 2026 (co-led by MGX among others; Claude Code run-rate over $2.5 billion; the seller’s own count of its largest buyers, read from the post on 19 September 2026: “Two years ago, a dozen customers spent over $1 million with us on an annualized basis. Today that number exceeds 500. Eight of the Fortune 10 are now Claude customers”; the same post relays an outside estimate, not a count of its own: “A recent analysis estimated that 4% of all GitHub public commits worldwide were being authored by Claude Code”, the analysis being SemiAnalysis’s, linked from the post). In the Sources
- Dwarkesh Patel, “Dylan Patel — The single biggest bottleneck to scaling AI compute,” Dwarkesh Podcast, YouTube, 13 March 2026 (Patel, the founder of SemiAnalysis, on TSMC’s 3-nanometer allocation: Apple was talking to other vendors because it was being squeezed out of TSMC, whose margins on AI chips are higher than on mobile; TSMC allocates more readily to CPU business, which it views as more stable; paraphrased, the published transcript being an edited one) In the Sources
- Dylan Patel, interviewed by Sarah Guo, “Chips, Neoclouds, and the Quest for AI Dominance with SemiAnalysis Founder and CEO Dylan Patel,” No Priors Ep. 127, 14 August 2025 (the bottleneck line at 28:58.6–29:03.5, Patel on camera; “if one thing was solved you could increase production another 20% and then something else would be the issue”; on the auto caption and both Whisper decodes of 25 September 2026, a slice from 28:00 held in book 1’s Quote audio checks folder of that day with word tables and its md5 in the manifest, Sourcing’s pool, INBOX 684 §118, the tape being this No Priors episode and not the a16z upload of 18 August 2025, xWRPXY8vLY4, which the row cited until rev 160 and whose caption never carries the word solved). In the Sources
- Doug O’Laughlin, “GPT-3 and the Writing on the Wall,” Fabricated Knowledge, 12 August 2020 (on GPT-3 and on OpenAI’s paper “Scaling Laws for Neural Language Models”: “I am not qualified to know if this is truly the way forward, or if it’s just another pit stop on the long road forward in AI research”; “The most interesting conclusion from an investor’s viewpoint is not GPT-3, but the writing on the wall”; “The writing on the wall is orders of magnitude more compute purchased”; the author had worked at an investment firm, Bowie Capital, by his newsletter’s own account, and joined SemiAnalysis in November 2024, four years after this post; held whole in the Library since Sourcing’s capture of 5 October 2026). In the Sources
- Data Center Dynamics, “xAI’s Memphis Supercluster Has Gone Live with Up to 100,000 Nvidia H100 GPUs,” 23 July 2024 (SemiAnalysis estimate of about 32,000 GPUs powered on 22 July 2024). In the Sources
And 41 more: search the Sources.
Related terms
From the glossary of The AI Compute Race.
How to cite this page
Dryburgh, Lee S. “SemiAnalysis.” In The AI Compute Race. DRYBURGH. https://dryburgh.com/vol1/glossary/semianalysis (accessed 2026-11-01).
BibTeX
@misc{dryburgh-semianalysis-2026,
author = {Dryburgh, Lee S.},
title = {{SemiAnalysis}},
howpublished = {In \emph{The AI Compute Race}, DRYBURGH},
year = {2026},
url = {https://dryburgh.com/vol1/glossary/semianalysis},
note = {Accessed 2026-11-01}
}