Physical Unclonable Function Patents: Top Companies & Trends 2026
- 22.5% concentration at the top. The five most active filers hold 265 of the 1,180 records in scope, but that still leaves a long tail of single- and few-filing entrants working the remaining claim space.
- Filings cooled after a 2021 high. Annual filings dropped from 123 in 2021 to 76 in 2024, a 38% pull-back over three years — though 2025-2026 counts are still filling in due to publication lag and should not be read as a continued decline.
- H04L dominates the claim map. 68.9% of all records touch digital information transmission (H04L), while measurement-focused G01R claims sit at just 4.2% — a gap worth checking before filing there.
Filing growth compares 2021 (123 records) with 2024 (76) — a three-year span. 2024 is the most recent year we treat as complete: publication lags filing by roughly 18 months, so 2025 onwards are still filling in and any growth rate that ends there would understate the field. Top-5 share is the combined record count of the five largest assignees divided by all 1,180 records in scope (CR5), not by the ranked leaders only.
What the physical unclonable function patent record shows
A physical unclonable function (PUF) exploits manufacturing-level randomness in a chip to produce a device-specific fingerprint that cannot be cloned even by the manufacturer. The patent record tracked here spans 1,180 published records from 2015 through mid-2026, drawn from filings that explicitly address reliability, uniqueness, helper data, temperature and voltage variation, modeling attacks, or silicon area constraints — the core engineering problems that separate a working PUF from a theoretical one. Helper data and error correction claims recur throughout the corpus because raw PUF outputs are noisy across temperature and voltage, and a cryptographic key generated from them has to be stable on every read.
Filing activity concentrates around digital transmission security and general data processing, with semiconductor-specific and memory-array implementations forming a smaller but persistent share. The mix suggests PUF technology is being claimed less as a standalone silicon primitive and more as a component inside broader authentication and key-management systems.
Filing trend and technology composition
The dataset covers 1,180 published records across receiving offices including the United States, EPO, China, WIPO, the United Kingdom and India, with IPC subclass tagging that lets several classes attach to a single record.
Filings peaked in 2017, then again around 2021
Annual filings hit an early peak of 127 in 2017, and after a lower stretch, climbed back to 123 in 2021 before falling to 76 in 2024 — a 38% decline over that three-year window. Counts for 2025 and 2026 are still incomplete because publication typically lags filing by around 18 months, so the most recent bars understate real activity.
H04L and G06F carry the bulk of the claim volume
H04L (digital information transmission) appears in 68.9% of the 1,180 records and G06F (electric digital data processing) in 33.5%, reflecting PUF's role inside larger authentication and key-management systems. Semiconductor-specific classes — H01L at 9.1% and G01R at 4.2% — carry far less claim density, which is where hardware-level differentiation is least contested.
Shares are the percentage of the 1,180 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Physical Unclonable Functions with Eureka
This page is one run against one query. Ask Eureka your own question about physical unclonable functions and every answer comes back with the patent numbers behind it.
Try EurekaMost-cited records and a representative claim
Reducing helper data size for physical unclonable function device (US20230094237A1, NXP B.V., 2023-03-30)
The application discloses a PUF device with an array of cells, where a reliable cell group detector identifies groups of cells reliable enough to produce stable outputs and stores only their addresses as helper data. This narrows the helper data footprint compared to schemes that store correction data for every cell, directly targeting the storage and bandwidth overhead that helper-data schemes typically carry.Filed by NXP, a name that recurs across memory-array and helper-data claims in this corpus.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20140108786A1 | Tamper-protected hardware and method for using same | 350 |
| 2 | US20120183135A1 | Reliable PUF value generation by pattern matching | 130 |
| 3 | US8525169B1 | Reliable physical unclonable function for device authentication | 129 |
| 4 | US20150195088A1 | PUF Authentication and Key-Exchange by Substring Matching | 126 |
| 5 | US20170063559A1 | Authentication system and device including physical unclonable function and threshold cryptography | 99 |
| 6 | US20140201851A1 | Method and apparatus for using dynamic voltage and frequency scaling with circuit-delay based integrated circ… | 92 |
| 7 | US20140042627A1 | Electronic structure containing a via array as a physical unclonable function | 92 |
| 8 | US20170149572A1 | Authenticatable device with reconfigurable physical unclonable functions | 89 |
| 9 | US20170200508A1 | PUF value generation using an Anti-fuse memory array | 81 |
| 10 | US20140225639A1 | Integrated circuit identification and dependability verification using ring oscillator based physical unclona… | 81 |
Citation counts favour older records simply because they have had longer to accumulate citations inside this searched corpus — read them as a signal of influence on the field, not as a marker of current importance.
Each row carries its publication number; clicking a row searches Eureka by that number.
Put your own technology through the same analysis
Eureka on the web
When you want the answer in the next five minutes.
The agent works the prompt against patents and technical literature, citing every source.
Run your analysis now →MCP server & REST API
When it has to run inside your own pipeline.
Patent search, landscape analysis and assignee resolution as MCP tools. Drop them into any agent framework, or call REST directly.
Browse MCP servers →Reading the concentration and the trend together
Two figures matter more than any single ranking: how much of the field the leaders actually hold, and whether the recent filing dip reflects real cooling or just publication lag.
Leadership is real but not dominant
The five most active assignees combined hold 265 of the 1,180 records in scope — 22.5% of the field. That leaves the large majority of filings spread across a long tail of companies and institutions, each holding a handful of records rather than a dominant position.
A real pull-back, not a collapse
Annual filings fell from 123 in 2021 to 76 in 2024, a 38% decline over three years. That is the last three-year span that can be read reliably; 2025 and 2026 figures are still filling in under the roughly 18-month publication lag and should not be read as confirming a continued slide.
Transmission-layer claims crowd out measurement claims
H04L (digital transmission) touches 68.9% of the 1,180 records, while G01R (electric and magnetic measurement) touches only 4.2%. The gap points to where PUF is being claimed as a system-level security feature far more often than as a raw measurement or characterization technique.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to physical unclonable functions, with the prior art for and against each one.
Who is filing, and where the activity has slowed
The ranking covers 100 companies returned by the data endpoint, not a curated top-50 or top-100 list. The leader holds 71 records; by fifth place that figure is 35, and by tenth place 26 — a steep early drop-off that flattens into a long tail.
One filer well ahead of the rest
The leading assignee holds 71 records, more than double the fifth-place holder's 35. That gap narrows quickly further down the ranking, where tenth place sits at 26.
Recent-year activity has gone quiet across the board
Among the assignees tracked for recent-year momentum, the most active logged just 1 filing in the latest year, with the rest at 0. That is consistent with the broader 2025-2026 undercount from publication lag rather than a sign that these firms have exited the space.
Co-filing is limited and concentrated
Only 10 co-assignee pairs appear across the corpus, with the strongest links tied to university-industry pairings and named-inventor partnerships rather than broad cross-company alliances.
| Assignee | Recent year | YoY |
|---|---|---|
| Taiwan Semiconductor Manufacturing Co., Ltd. (TSMC) | 1 | — |
| Merck Patent GmbH | 0 | — |
| Lexmark International, Inc. | 0 | — |
| International Business Machines Corporation (IBM) | 0 | — |
| Intrinsic ID B.V. | 0 | — |
| STC.UNM | 0 | — |
| Macronix International Co., Ltd. | 0 | — |
| Analog Devices, Inc. | 0 | — |
Where to take this analysis
The figures above establish the shape of the field. Turning that into a filing or freedom-to-operate decision means going record by record against a specific claim set.
Check freedom-to-operate against the cited leaders
The most-cited records in this corpus, including US20140108786A1 and US20120183135A1, define much of the prior art that later helper-data and reliability claims have to work around.
Search these records in EurekaTrack the under-claimed branches before they fill in
Measurement-based characterization and silicon-area-constrained layouts carry far lower filing density than transmission-layer claims — a narrower window to stake a first claim.
Monitor these branches in EurekaCommon questions about PUF patents
In patent claims, a physical unclonable function is typically defined as a hardware circuit or array of cells that produces a device-specific output derived from uncontrollable manufacturing variation, used to generate a cryptographic key or authentication response. Claims in this corpus most often cover the reliability mechanisms around that core idea — helper data generation, error correction across temperature and voltage swings, and reliable-cell selection — rather than the raw physical effect itself. That is why so many filings sit in H04L and G06F rather than purely semiconductor-device classes.
The ranked leaders in this dataset span semiconductor manufacturers, memory and printing technology firms, university research offices and specialist security companies, with the top-ranked assignee holding 71 of the 1,180 records tracked. The top five combined hold 265 records, 22.5% of the field, but the ranking drops off quickly after that into a long tail of smaller filers. No single company controls a majority of the claim space.
Filings peaked at 127 in 2017 and again reached 123 in 2021, before falling to 76 in 2024, a 38% decline over that three-year span. Figures for 2025 and 2026 are still incomplete because patent publication lags filing by roughly 18 months, so it is too early to read the most recent years as confirming a continued slowdown. The honest read is a real pull-back through 2024 with an open question beyond that.
Digital information transmission (H04L) appears in 68.9% of the 1,180 records in scope, and general electric data processing (G06F) in 33.5%, reflecting how often PUF is claimed as part of a larger authentication or key-management system rather than as a standalone hardware primitive. Ciphering and secret communication (G09C) and static memory (G11C) each cover more than 13% of records. Measurement-focused claims under G01R are comparatively rare at 4.2%, suggesting less-crowded claim space there.
US20230094237A1, filed by NXP, claims a method of reducing helper data size by detecting and storing only the addresses of reliable cell groups within a PUF array, rather than correction data for every cell. Anyone building a helper-data scheme for area- or bandwidth-constrained devices needs to check their reliable-cell selection and storage approach against this claim set. Workable alternatives generally involve different criteria for what counts as a reliable group or different data structures for storing the addresses, rather than avoiding reliable-cell filtering altogether.
Research Physical Unclonable Functions in depth with Eureka
Go past this page: query the whole physical unclonable functions corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Disclaimer. This page is generated from Patsnap Eureka data drawn from a limited snapshot of global patent and scientific-literature records, and is provided for general information and reference only.
Patent data carries inherent limitations: recent filings (typically the most recent 18–24 months) are under-counted due to standard publication lag; counts may be reported at either a patent-family or a patent-record basis and are not always directly comparable; classification, applicant-name, and citation data may contain errors, duplicates, or omissions; and the underlying search query defines and constrains the scope shown. As a result, the analysis may be incomplete or inaccurate and may not reflect the full technology landscape.
Nothing on this page constitutes an exhaustive prior-art, novelty, freedom-to-operate, or validity search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. Any patent, commercial, or strategic decision should be verified independently and reviewed with qualified patent, legal, and domain professionals. Patsnap makes no warranties, express or implied, as to the accuracy, completeness, or fitness for any particular purpose of the information presented.
Machine translation. Assignee and organisation names originally recorded in Chinese, Japanese or Korean have been rendered into English by an AI translation step so that the tables stay readable. These renderings are best-effort and may not match a company’s registered English name; the original name is what the underlying patent record carries, and it is what any Eureka query launched from this page uses.