Method · Least certain

Where this record is least certain.

Everything this record knows it is missing, with the reason it is missing and what recording it would unblock. A tracker that shows only what it has is telling you half of what it knows.

Every figure links to its publisher What changed, and when →

If you know a publisher we are not reading, that is the most useful thing you can send us.

Please do not send a figure. A number without a publisher and a date cannot go on this site, whoever it comes from, and sitting next to figures that have both it would look exactly like one of them. Send the source and the record does the rest.

record@siliconalmanac.com →

24 named gaps

Grouped by why each one is open, in an order this record chose: the further down the page, the closer the gap is to being closable by work rather than by somebody else changing their mind. It is an order, not a score, and nothing on this page is rated.

Nobody publishes it 2no-source

No publisher appears to produce this at all. The gap is in the world, not in the collection.

The price of the nitrogen a fab uses, which is the gas it uses most of

Nobody appears to publish it. The obvious series to reach for is the USGS nitrogen summary, and it measures a different product. Labelling ammonia as fab nitrogen would be a wrong number with a real publisher's name on it.

Known: Fab nitrogen is separated from air in units on or beside the site, so what it costs is mostly the electricity those units draw. The U.S. Geological Survey's summary titled Nitrogen is "Nitrogen (fixed), ammonia": fertiliser feedstock, priced per short ton at the Gulf Coast, and not this gas.
Would unblock: nothing directly; the nearest published proxy is industrial electricity, which has its own entry
Source: https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-nitrogen.pdf

Worldwide semiconductor equipment billings for any FOURTH quarter, as a quarterly figure

The annual total is published and the three quarters are published, so Q4 is one subtraction away, and it is not taken. SEMI restates these figures, so a residual would silently mix a restated annual with unrestated quarters, and a restatement is the subject this record exists to expose. The detailed WWSEMS report, which does carry every quarter, is SEMI's paid Equipment Market Data Subscription. The Semiconductor Equipment Association of Japan publishes the same tables and requires written permission to reprint, on a condition of non-commercial use.

Known: SEMI publishes the Worldwide Semiconductor Equipment Market Statistics headline for Q1, Q2 and Q3 as a free press release, about nine weeks after each quarter ends. Measured over all 14 billings releases it has issued since Q1 2023: there is no fourth-quarter release in the whole run. The year closes with an annual release in about April which states the full-year total and segment growth and names no fourth quarter. The quarterly series this record holds therefore runs Q1 to Q3 of each year.
Would unblock: a complete quarterly billings series, and any claim about a peak that falls in a fourth quarter
Source: https://www.semi.org/en/products-services/market-data

It exists and it is not free 5not-free

A source exists and is not free or not public, so it is out of scope permanently.

A comparable industrial water price, per country, for the countries this record's fab sites are in

The survey is a paid product and this record buys no data, ever. Residential tariffs from individual utilities are not an industrial price, and averaging them into a country figure would produce a number no publisher printed. Water is a real input cost for a fab, and the honest state of the free public record is that nobody publishes it comparably.

Known: Global Water Intelligence publishes a Global Water Tariff Survey covering hundreds of cities across most countries, sold as a report. IBNET holds utility-level tariffs with uneven coverage and is largely residential.
Would unblock: the water half of any page about what a fab site costs to run
Source: https://www.globalwaterintel.com/global-water-tariff-survey

A forecast of semiconductor sales by end use (computers, communications, automotive, industrial, consumer, government), for a stagger chart beside the published shares

A stagger chart needs the same target forecast repeatedly and then its result. The results exist and are on the map; the forecasts are paid products, and this record buys no data. Deriving an end-use forecast by applying a past share to a WSTS total would be a number no publisher printed.

Known: The Semiconductor Industry Association publishes each year's end-use shares after the fact, citing the WSTS End-Use Survey. WSTS publishes its forecasts free by product category and by region, never by end use. Forecasts by end market are sold as research reports.
Would unblock: a stagger chart of end-use shares on the map
Source: https://www.wsts.org/61/Forecasts

A price for neon, krypton and xenon, the rare gases chip lithography and etch use

The only prices are paid products, and this record buys no data. Dividing a customs import value by an import weight would produce a unit value no publisher printed, for a gas whose grades differ by orders of magnitude in price.

Known: The U.S. Geological Survey's 2026 summary "Helium and rare gases" gives US volumes for all four and puts semiconductors at 32% of US neon use, 67% of krypton and 36% of xenon. It gives a base price for Grade-A helium, about $12 per cubic meter in 2025, and then says: "Price data for rare gases were unavailable." Every neon price found on 16 September 2026 was a commercial market report sold by a research firm.
Would unblock: a price beside the rare-gas volumes on /materials/
Source: https://pubs.usgs.gov/periodicals/mcs2026/mcs2026-helium.pdf

Knometa Research's Global Wafer Capacity, more than 400 fabs with capacity, wafer starts and utilisation. The 2024 edition is the LAST: the publisher states there will be no further issues.

Paid. A figure from it could inform what to look for in public sources and could never appear on this site, so buying it would be research rather than supply. Recorded because the publisher states the 2024 edition is the last, so this database cannot be a check on any year after 2024.

Known: Final edition on sale at US$2,750 single user, US$4,200 enterprise, read from knometa.com on 17 August 2026.
Would unblock: a check on whether this record's free-sourced picture is right
Source: https://knometa.com/services/reports/

Korean semiconductor export figures at source

Not a licensing problem but an access one. UN Comtrade covers a narrower commodity code months later, and its licence forbids automated download and publication, so it is not a substitute: a Comtrade series ran here until 22 September 2026 and was withdrawn for that reason.

Known: data.go.kr publishes them; all three membership types require Korean identity verification or a Korean business registration
Would unblock: a faster Korean read on the indicators page
Source: https://www.data.go.kr/

The publisher does not allow automated reading 2publisher-disallows-automation

Free for a human to read, and the publisher has instructed automated clients not to fetch it. Different from not-free, where the material itself costs money, and from permission-required, which is about republishing material already in hand. Here the objection is to the FETCH, so no collector runs against it and no second route is used to get around the instruction. A human reading the same page in a browser is not an automated client and is not covered by it.

UMC's quarterly report PDFs, which carry a per-fab capacity table in thousands of 8-inch equivalent wafers - the most granular first-party capacity disclosure any foundry makes, quarterly.

The publisher has told automated clients not to. robots.txt disallows /upload/ for every user-agent, and that is where every quarterly report lives. This record does not fetch what a publisher has asked automated clients not to fetch, and it does not route around the instruction by another means. A HUMAN opening the PDF in a browser is not an automated client and is not covered by it, so the figures can be read and hand-entered the way every other entry in fab-capacity.json was - the file is hand-curated by design. What cannot happen is a collector. [Re-filed 18 August 2026 from method-cannot-reach, which was the wrong shelf: the method can reach these files perfectly well and the publisher said not to. The distinction got its own status when a second case appeared, see youtube-channel-feeds.]

Known: The files exist and their URLs are stable: umc.com/upload/media/08_Investors/Financials/Quarterly_Results/... UMC's robots.txt says `User-agent: *` and `Disallow: /upload/`.
Would unblock: Per-fab quarterly capacity for a whole foundry, on a stated 8-inch-equivalent basis, with a history going back to 2000.
Source: https://www.umc.com/en/Download/quarterly_results

Uploads from nine YouTube channels this record used to collect: Asianometry, TechTechPotato, Ben Eater, Applied Science, Branch Education, ASML official, imec, der8auer and High Yield.

TWO INDEPENDENT REFUSALS, and the second one is the surprise. First, the feed path is disallowed to every automated client, and this record does not fetch what a publisher has told automated clients not to fetch. Second, the sanctioned alternative closes too: the YouTube API Services developer policies allow an API client to store non-authorized data for no longer than 30 calendar days, after which it must delete or refresh it. This archive's central invariant is that it never shrinks and holds items back to 1999, so the compliant route and the archive contradict each other directly. Neither refusal is a technical fault and no user-agent change cures either.

Known: Each was collected from https://www.youtube.com/feeds/videos.xml?channel_id=... Read 18 August 2026, youtube.com/robots.txt contains, under `User-agent: *`, the line `Disallow: /feeds/videos.xml`. Six of the nine had begun returning 404 or 500 and three still answered, which is inconsistent enforcement on YouTube's side rather than a difference between the channels.
Would unblock: nothing automatically. A human subscribed to these channels in a browser is not an automated client, so an individual video worth recording can be hand-entered the way fab-capacity.json already is. Where a creator publishes their own newsletter the record takes that instead: More Than Moore was added on 18 August 2026 for exactly this reason, and Asianometry's Substack was already held.
Source: https://www.youtube.com/robots.txt

Refused by a server, permitted or unstated by the publisher 2blocked-not-disallowed

The publisher's server refuses an automated fetch while the publisher has NOT instructed automated clients to stay away: either their robots.txt permits it, or their server will not serve a robots.txt at all so no instruction can be read. Different from publisher-disallows-automation, where the publisher said no and this record obeys. Here nobody said no and the refusal is a bot-detection heuristic. This record does not disguise itself to get past one: sending a browser's user-agent in order to be served is the same act whether or not a rule was broken, and a record whose whole claim is that it says where its figures came from cannot also be lying about who is asking. The route left open is to ask the publisher.

Micron's investor-relations press releases, as a feed, at investors.micron.com/rss/pressrelease.aspx

THE PUBLISHER'S STATED POLICY AND THEIR SERVER DISAGREE. robots.txt allows every user-agent and asks for ten seconds between requests; the feed answers HTTP 403 Forbidden to a request that honours both. Measured 0 successes in 133 attempts since the collector started, from two separate networks, on HTTP/2 and HTTP/1.1. This record will not send a browser's user-agent to get around a bot rule, because looking less like a robot in order to be served is the same act whether or not a rule was broken. Micron's announcements do reach this archive, through the trade press; what is missing is the primary link, not the fact. Their crawl delay of 10 seconds is honoured. Until 1 September 2026 each run sent three requests inside about three seconds, because a failed fetch was retried twice with no spacing.

Known: Their robots.txt is served to this record's user-agent and reads `User-agent: *`, `Crawl-delay: 10`, `Allow: /`. The site runs on q4web.com, Q4 Inc's investor-relations platform.
Would unblock: Micron press releases citable to Micron rather than to a publication reporting on Micron.
Source: https://investors.micron.com/rss/pressrelease.aspx

The investor-relations and newsroom feeds of Applied Materials, Texas Instruments and STMicroelectronics

THIS RECORD CANNOT READ THEIR INSTRUCTION, SO IT CANNOT CLAIM TO BE FOLLOWING IT. A request for robots.txt times out from two separate networks, on HTTP/2 and on HTTP/1.1, so there is no way to learn whether automated access is permitted. That is why these sit here and not under publisher-disallows-automation: nobody has disallowed anything that this record has been able to read. Measured 0 successes in 133 attempts each. Over HTTP/2 the connection is opened and the stream is killed with INTERNAL_ERROR; over HTTP/1.1 the feed path returns 403 while robots.txt returns nothing. Each of these three costs 141 seconds of every collector run, 45 seconds times three attempts, which is most of the run's wall-clock time.

Known: ir.appliedmaterials.com and investor.ti.com resolve through gcs-web.com to Akamai; www.st.com resolves to st.com.edgekey.net, also Akamai. All three carry a valid feed URL that a browser loads.
Would unblock: Three of the four dead entries in the company-newsroom bucket, which runs at three working feeds out of seven.
Source: https://ir.appliedmaterials.com/rss/news-releases.xml

The licence does not clearly permit it 1licence-unclear

The source has been read and its own statements about reuse CONTRADICT each other, so nothing from it is published until the publisher settles it. Different from not-free: the material is free to obtain and the question is whether it may be republished.

The IEEE International Roadmap for Devices and Systems (IRDS). Its More Moore chapter publishes the Logic Core Device Technology Roadmap as a SPREADSHEET, not only as a PDF, and the tables were read successfully on 17 August 2026: 2024IRDS_MM_Tables.xlsx and the 2022 equivalent both parse, and both carry a row named "Logic industry 'Node Range' Labeling" that maps each year of production to the label the industry uses for it.

This record republishes figures. A source whose own file says permission is required for republishing is not publishable on the strength of a friendlier sentence on its website, and the two cannot both be relied on. Nothing from these tables is reproduced anywhere on this site, including in this entry: what follows describes the SHAPE of the roadmap without copying it.

Known: TWO STATEMENTS FROM THE SAME PUBLISHER, AND THEY DISAGREE. The website says: "The IRDS roadmap reports can be used freely and with permission by giving proper credit to IEEE and the IRDS. IRDS images, data, and content can be shared for the greater good of the industry and research community." The spreadsheet's own index sheet says: "© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works." Note that even the website's sentence says "freely AND WITH PERMISSION", which is not a permission.
Would unblock: A SECOND PUBLISHER for node bands, and something no surface anywhere presents: the roadmap as a revision record. Comparing the 2022 and 2024 editions of one row shows the roadmap moving in three separate ways - the step between production years shortened, one step moved a year earlier while being renamed, and the naming convention changed from nanometre-equivalent to angstrom-equivalent, so the same physical target reads as a different label in the two editions. That is the stagger-chart pattern in a place nobody looks for it, and ten editions exist.
Source: https://irds.ieee.org/editions/

It needs written permission 1permission-required

Free to obtain and read, and the publisher's own licence prohibits republishing it without written permission. Different from not-free, where the material itself costs money, and from licence-unclear, where the publisher's statements contradict each other. Here the statement is clear and it is no.

UN Comtrade: trade flows by HS code, reporter and partner. HS 8542 is electronic integrated circuits, 8486 is semiconductor manufacturing machinery, 3818 is doped wafers. Probed directly on 17 August 2026 rather than read about.

The licence prohibits, by name, the two things this record would do with it: AUTOMATED DOWNLOADING and PUBLICATION. An API key would not change that - a key raises a rate limit, it does not grant a licence - so registering would only add a credential to an activity the terms forbid. Nothing from UN Comtrade is collected or published here, and the endpoint was probed by hand a handful of times to establish exactly this.

Known: TWO FACTS, AND THE SECOND OVERRIDES THE FIRST. The free endpoint works with no key: comtradeapi.un.org/public/v1/preview returned Netherlands imports of HS 8542 for 2023 and 2022 without any credential. The rate limit is severe and exposed in no header - four calls in quick succession gave one 200 and three 429s, one call after a 25-second pause succeeded. Rows carry estimation flags: both read had isReported=false and isNetWgtEstimated=true, so UN Comtrade compiled or estimated them rather than the reporter filing them. The UN Comtrade License Agreement then says, of the material: "Except as may otherwise be expressly permitted by applicable law, any copying, automated browsing or downloading, redistribution, publication, or commercial exploitation of any material contained on or otherwise made available to you on United Nations COMTRADE is strictly prohibited without the prior written permission of the United Nations."
Would unblock: Trade flow as a public series, which is the only evidence that would show the industry's physical movement between countries. Worth one written request to comtrade@un.org, on the same pattern as the IRDS letter.
Source: https://comtradeapi.un.org/public/v1/preview/C/A/HS

Published, and out of this method's reach 2method-cannot-reach

The material is collected and published, but the method used to index it cannot address this subject without guessing, so it is not indexed rather than indexed wrongly.

Archive items about Arm Holdings, indexed on the company page

Company pages list an item when a headline contains one of a declared set of strings, matched literally and case-sensitively at word edges. 'Arm' is three characters, and 'arm', 'arms' and 'armed' are ordinary words that a Title Case headline capitalises, so the short name cannot be used. 'Arm Holdings' would be safe and would match almost nothing, because coverage writes 'Arm' alone. A list built from either choice would read as coverage while being either wrong or nearly empty, so no list is published. Same rule as everywhere else here: a gap is never closed by guessing.

Known: The items are collected and are already published on /archive/, reachable by source and by free-text search.
Would unblock: An indexed item list on the Arm company page. The page itself still carries the catalogue facts, its fab sites and its links.
Source: https://siliconalmanac.com/archive/

WSTS press releases older than the eleven its archive currently lists

The archive is the only index WSTS publishes and it is a moving window. The media ids do not follow from the date, so a URL for a release that has scrolled off cannot be constructed - only found somewhere that is not the publisher, which is not this record's source of truth.

Known: The archive at /76/PRESS-ARCHIVE listed eleven items on 17 August 2026, the oldest being the Spring 2023 forecast round. Each PDF sits at a path containing an opaque numeric media id.
Would unblock: a 2023 target line with three vintages and a result, and the result marker on the 2024 line
Source: https://www.wsts.org/76/PRESS-ARCHIVE

Believed to exist and not located 3believed-to-exist

A publication of this kind is known to be issued on a cadence, but this specific one has not been located.

A current US import price index for electronic integrated circuits, HS 8542

A price index that stopped in 2022 would sit beside 2026 figures and read as current, so the stopped series is not published at all. FRED ending a series does not show that BLS stopped compiling it: BLS may publish it under another code or table, and that has not been checked.

Known: The Bureau of Labor Statistics import price index for HS 8542 is on FRED as IP8542, index 2000=100, and its last observation there is December 2022. The export index for the same code, ID8542, runs to August 2026 and this record holds it. Read off FRED on 23 September 2026.
Would unblock: the import side of the US price read, beside the export index and the producer price index this record holds
Source: https://fred.stlouisfed.org/series/IP8542

The WSTS full-year 2024 result

Deriving 2024 by dividing the 2025 result by its growth rate would produce a number WSTS never printed, presented as if it had. The 2024 target therefore carries a forecast and no result.

Known: The 2025 result release states 26% growth to USD 795.6 billion, which implies a 2024 figure
Would unblock: the result marker on the WSTS 2024 line
Source: https://www.wsts.org/76/103/Global-Semiconductor-Market-grows-26-in-2025-to-796B

WSTS quarterly releases other than Q2 2025

WSTS's public press archive holds only the most recent releases - eleven items in August 2026, the oldest a Spring 2023 forecast - and there is no index of past quarterlies. A URL cannot be guessed: the PDF paths carry an opaque numeric media id (7740 for Q2 2026, 7175 for Q2 2025) that does not follow from the date.

Known: WSTS issues quarterly releases that update the standing forecast. Two are now in the record: Q2 2025, which moved the 2026 target from 760.7 to 800, and Q2 2026, which moved 2026 to 1,655 and 2027 to approximately 2.1 trillion.
Would unblock: denser WSTS lines on every target
Source: https://www.wsts.org/61/Forecasts

Read, and a required field is missing 5known-not-recorded

The figure has been read and the source is known, but a required field is missing and will not be guessed.

300mm fab equipment spending for target 2024

A single point is a first forecast, not a stagger. It is held out rather than drawn as a line of one.

Known: The June 2023 release gives US$82 billion for 2024
Would unblock: a 2024 row in the SEMI 300mm matrix
Source: https://www.semi.org/en/news-media-press-releases/semi-press-releases/global-300mm-fab-equipment-spending-forecast-to-reach-record-$119-billion-in-2026-semi-reports

Company logos on the company pages

A logo is a trade mark, and the free-and-public rule that governs the data does not extend to marks. More immediately: every number on this site links to the body that published it, so a logo taken from a search result would be the only asset here with no recorded origin. Files are therefore sourced from each company's own press kit or brand centre, one at a time, or the company renders as text with the reason published.

Known: The machinery is built and held: a provenance record per logo, a check that refuses any logo file without an entry, and pages that render the company name as text until an entry is complete.
Would unblock: Faster recognition on a dense page. A reader recognizes a mark before they read a name.
Source: https://siliconalmanac.com/companies/

One worldwide series of industrial electricity prices, on one basis

Stitching several publishers into one map or one colour scale would assert a comparison none of them made: consumption band, tax treatment, currency and period all differ between them. This record would rather show each publisher's own panel, named and dated, than invent a world series. It would also be wrong in a second way that no caption fixes: a national average is not what a large fab pays, because industrial consumers of that size negotiate contracts.

Known: Eurostat publishes electricity prices for non-household consumers by consumption band, twice a year, free and redistributable, for the EU and several neighbours (nrg_pc_205). The US EIA publishes industrial-sector prices by state in the public domain. The IEA publishes a global set under licence.
Would unblock: a per-country electricity panel beside the fab map, one publisher per panel
Source: https://ec.europa.eu/eurostat/databrowser/view/nrg_pc_205/default/table?lang=en

Process nodes for eighteen fab sites, removed from the record on 21 August 2026 for want of a publisher

Writing the twelve supported figures back into a bare list would rebuild the exact fault that took a day to find. The per-site capacity record already holds the right shape for a per-site figure (the statement, its wording, its publisher, its date), and nodes need the same before any of the research is written back.

Known: Sources were chased for all eighteen. Twelve of the values are supported by a publisher naming that site, among them BAE Systems Nashua (100nm and 70nm PHEMT, 180nm and 140nm GaN), Microchip Gresham (0.11 to 0.5 microns), Nuvoton's ex-Winbond Fab 2 (0.35um to 1.0um), FBK Trento (500 nm minimum line width) and Qorvo Richardson (90 nm pHEMT, 0.25 micron GaN). Eleven are unsupported or contradicted: Qorvo Hillsboro is recorded at 500 where TriQuint published 0.15 um, and Denso Iwate at 1500 and 350 where the only node its operator publishes is 0.18 micron.
Would unblock: the node bands on /nodes/ for eighteen more sites, and a citable figure page for each
Source: https://siliconalmanac.com/nodes/

TSMC's own company-wide capacity figure, from its Fab Capacity page: "Annual capacity of the manufacturing facilities managed by TSMC and its subsidiaries exceeded 17 million 12-inch equivalent wafers in 2025." The page then names the facilities counted, including JASM as a majority-owned subsidiary.

The page carries no date and no dated meta field; the only timestamp available is the server's Last-Modified, which is when the page was served and not when the sentence was written. Every figure in this record is positioned by the date its publisher issued it, and a guessed date puts a real number on the wrong day - the same refusal the WSTS Spring 2024 entry makes. The dated version of this sentence is almost certainly in TSMC's annual report, which is where to look next.

Known: The figure, the scope, the bound - "exceeded" is a lower bound - and a unit this record does not yet hold: 12-INCH EQUIVALENT WAFERS PER YEAR. That is a fourth wafer basis, distinct from the 300mm ACTUAL basis already in use, and an annual period where every other capacity figure here is monthly or weekly.
Would unblock: A COMPANY-WIDE capacity figure from the company itself, which is the one kind fab-capacity.json refuses to compute: that file holds per-fab statements and never adds them up. A publisher's own total is the only honest way to have one, and TSMC restates this sentence every year, which makes it a revision series rather than a fact.
Source: https://www.tsmc.com/english/dedicatedFoundry/manufacturing/fab_capacity

Held back on purpose, for a stated reason 1held-by-decision

Possible to build, and deliberately not built, by a decision this record states with its reason. Different from built-not-run, where code exists and has not met the live endpoint: here nothing is built, on purpose, and the entry says why.

Schedule revisions: a promised milestone date moving, such as a fab's start of production

It is the table about named companies. The neutral visual and verbal language is being proved first on statistical-agency and trade-body figures, where no party is embarrassed by the chart.

Known: The schema is decided and written down. Company capital-spending and revenue guidance is the other half of this table, and on 22 September 2026 it was decided that guidance is recorded as vintages, from each company's own earnings release, first-party statements only.
Would unblock: the third table

Several of these could be closed by guessing and are deliberately left open. A guessed date puts a real number on the wrong day, which is a worse error than an absent point.

Counted, rather than written up

The list above is curated, so it keeps up with attention rather than with the data. This one is not: every guard in this record produces a held-back pile as a side effect, and until they were counted together they were visible to nobody.

Held back	Count	Out of	File
Per-fab capacity figures held back Their only source is secondary - Wikipedia or trade press - and they state no wafer basis. They sat on the fab rows as a bare number until 17 August 2026, rendered nowhere, one surface away from being published as fact.	12	of 22 figures the record has touched	fab-capacity.json
Published capacity sentences read and refused First-party publishers said something about capacity that cannot honestly become a figure: an INCREMENT with no level behind it, a magnitude like "tens of millions" where the difference between 20 and 90 million is the whole fact. Kept with the reason so the next reader of the same release finds the decision already made.	2	of 24 capacity statements this record has read	fab-capacity.json
Capacity figures recorded and rendered on no page The fab is on the map and its operator is not a company this record covers, so there is no company page for the figure to sit on. The figure is right and nobody can see it, which is a different state from being held back and is counted separately for that reason.	2	of 10 figures recorded	fab-capacity.json
Node values that cannot be placed in a published band A band is a statement about nanometres. These are materials, memory generations, packaging technologies, named processes like N4 and Intel 18A, one lithography tool, and bare numbers with no unit. Translating a named process into a figure would put this record's inference inside a publisher's definition.	69	of 88 distinct node values in the fab rows	node-vocabulary.json
Fab sites that appear on no company page Their operator is not a company in the universe: subsidiaries, research institutes, and firms this record does not cover. The site is on the map; it just has no company page to sit under.	44	of 191 sites on the fab map	fab-operators.json
Feeds no human has confirmed by hand Nobody has yet fetched them by hand and seen a real feed come back. This is NOT a claim that they are broken - several are among the archive's largest sources, because the production server reaches hosts a sandbox cannot. It is a claim about what has been checked.	24	of 45 feeds configured	sources.json
Sources dropped, with the reason kept Dead URLs, bot walls and paywalls. They stay listed with the reason because a source removed silently is a gap nobody can audit.	26	of 71 sources ever considered	sources.json
Company logos refused The company's own terms restrict third-party use, or could not be found. Those companies render as text, and the whole set is held shut anyway pending an IP consultation.	2	of 2 logos examined	company-logos.json
Companies with no safe string to match on Every candidate string is a common English word, so matching on it would fill the page with items about something else. Recorded as a gap rather than papered over with a bad match.	1	of 177 companies	company-match.json
Gaps written up by hand, with what would unblock them The curated register: a missing publication date, a source that is not free, a collector built and not yet run. Each names what exists, what is missing, why, and what it would unblock.	24	—	record-gaps.json

A count of zero renders as zero. A row that disappeared would look like progress and would usually mean a renamed field. These are this record's own counts of its own files, not figures any publisher issued, which is why the last column names a file rather than a source, and why the table carries the date it was counted: 2026-10-04.

Announced elsewhere, not held here

Sites a publisher has said are being built, for which this record holds no row yet. It is produced by a rule rather than by judgement: a headline naming construction, and an operator this record already knows. An entry leaves this list when the row is added, and nothing on it has been assessed for importance.

Loading the current list…

This list is counted on the server every six hours, so it is as current as the last collector run rather than as the last time this page was built. It is this record's count of its own coverage, not a figure any publisher issued. A site announced before this archive began collecting on 13 August 2026 cannot appear here, because the rule reads headlines and the headlines from before that date were never collected.

What this page is not

Not a confidence score. Nothing here is rated, ranked by quality, or given a percentage. The grouping says why a gap is open and stops there.
Not a list of things that are wrong. Everything published on this site is what a named publisher published. These are the things that are absent, which is a different claim.
Not a request for numbers. See the box at the top.

← Back to the overview