Reliability Physics & Wearout Patents: Who Leads, Trends 2026
- Filing has cooled from its 2017 peak of 58 records but the 2021–2024 decline is a modest -10%, not a collapse — and 2024 is the last year publication lag lets us call complete.
- The top 5 assignees hold 37.5% of all 785 records while the top 10 combined reach 51.8% — concentration is real but a long tail of single- and few-filing entrants still fills out the rest of the ranking.
- H01L dominates at 53.8% of records but G01R measurement claims (27.8%) and the newer H10P and H10W subclasses point to test-structure and packaging-level reliability work as the more open ground.
Filing growth compares 2021 (20 records) with 2024 (18) — 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 785 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This landscape tracks patent families addressing bias temperature instability, hot carrier injection, and time dependent dielectric breakdown, paired with the acceleration and extrapolation methods — activation energy, voltage acceleration, electromigration lifetime, and stress migration — that turn stress-test data into qualified lifetime claims. The scope is deliberately narrow: it is not general semiconductor reliability, but the specific intersection of wearout mechanism and the quantitative method used to project device life from accelerated stress.
785 records published between 2015 and mid-2026 sit in scope, concentrated heavily in device structure (H01L) and electrical measurement (G01R) classifications, with a meaningful secondary cluster in digital processing (G06F) suggesting reliability prediction is increasingly handled as a computation and modeling problem, not only a materials or process one.
Let an AI agent run this analysis on your own technology
Pick a task. Every answer cites the patents behind it.
Filing trend and technology composition
Two views of the same 785 records: how filing activity has moved year over year, and which IPC subclasses carry the claim density.
Filing trend: a 2017 peak, then a plateau
Filings peaked at 58 records in 2017 and have since settled to a lower plateau; the 2021 to 2024 span shows a -10% change (20 to 18 records), a mild pullback rather than a sustained decline. 2025 and 2026 figures are still filling in under normal publication lag and should not be read as a drop-off.
IPC composition: device structure leads, measurement follows
H01L (semiconductor devices) appears in 53.8% of the 785 records, with G01R (electric and magnetic measurement) at 27.8% and H10P at 21.4% forming the next tier. G06F's 16.2% share indicates a sizable body of work treats reliability prediction as a data-processing and modeling task alongside the physical test structure itself.
Shares are the percentage of the 785 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Reliability Physics and Wearout Mechanisms with Eureka
This page is one run against one query. Ask Eureka your own question about reliability physics and wearout mechanisms and every answer comes back with the patent numbers behind it.
Try EurekaA representative test-structure claim
Test structure, method and circuit for simultaneously testing time dependent dielectric breakdown and electromigration or stress migration
Test structures for simultaneously testing for electromigration or stress migration fails and time dependent dielectric breakdown fails in integrated circuits, using four test structures arranged as a bridge balance circuit. The electromigration or stress migration portions use via chains of wire segments connected in series by conductive vias across at least two adjacent wiring levels. The time dependent dielectric breakdown portions use digitized wire structures in one of those same wiring levels, allowing both failure modes to be characterized from a shared, area-efficient test vehicle.Filed by International Business Machines Corporation, published 2013-02-14.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20020098627A1 | Surface preparation prior to deposition | 910 |
| 2 | US8242028B1 | UV treatment of etch stop and hard mask films for selectivity and hermeticity enhancement | 512 |
| 3 | US20060292845A1 | Processing substrates using site-isolated processing | 471 |
| 4 | US20070082508A1 | Methods for discretized processing and process sequence integration of regions of a substrate | 386 |
| 5 | US20040121620A1 | Surface preparation prior to deposition | 386 |
| 6 | US6303963B1 | Electro-optical device and semiconductor circuit | 374 |
| 7 | US20080156769A1 | Advanced mixing system for integrated tool having site-isolated reactors | 352 |
| 8 | US20070199510A1 | Systems and methods for sealing in site-isolated reactors | 348 |
| 9 | US6613695B2 | Surface preparation prior to deposition | 317 |
| 10 | US20080106925A1 | Correlated electron memory | 284 |
Citation counts favor older filings that have had more time to accumulate citations within this corpus; treat them as a signal of influence rather than current relevance.
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 →What the data means for filing strategy
Three patterns stand out once concentration, filing trend and classification data are read together.
Leadership is real but not exclusionary
The leading assignee holds 90 records, with fifth place at 34 and tenth at 18 — a steep drop after the leader but a broad plateau below it. Combined with the top 10's 51.8% share, this points to a field where one or two large filers set the pace but nearly half of all activity comes from outside the ranked leaders entirely.
A plateau after a 2017 peak, not a retreat
Filing peaked at 58 records in 2017 and has settled lower since, with the last fully comparable span (2021 to 2024) showing a modest -10% change. Because publication trails filing by roughly 18 months, 2025 and 2026 counts will rise as records catch up and should not be read as evidence of a cooling field.
Reliability prediction is becoming a computation problem
Alongside the expected H01L device-structure and G01R measurement classes, a sixth of records also carry G06F digital-processing classification. That overlap suggests lifetime extrapolation and acceleration-factor modeling are increasingly claimed as algorithmic methods, not just as physical test structures.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to reliability physics and wearout mechanisms, with the prior art for and against each one.
Who is filing, and where the gate sits
The ranking spans 100 companies across 785 records, from a clear leader down through a long tail of narrower, more recent entrants — several of them working test-structure and in-line measurement angles rather than core device physics.
A wide lead built on breadth, not a single mechanism
The top-ranked assignee's 90 records outpace the fifth-place filer's 34 by a wide margin, consistent with a portfolio built across multiple wearout mechanisms and test methodologies rather than a single narrow claim family.
A crowded middle tier below the top five
The gap from fifth (34 records) to tenth (18 records) is gradual rather than a cliff, indicating a genuinely competitive mid-tier of assignees rather than a sharp leader/follower split.
A small but tight cluster of joint filings
Ten co-assignee pairs appear in the data, with the strongest links repeating across the same three organizations — a sign of established joint development relationships around dielectric and electromigration test methods rather than one-off collaborations.
| Assignee | Recent year | YoY |
|---|---|---|
| International Business Machines Corporation | 0 | — |
| Intermolecular Inc | 0 | — |
| Taiwan Semiconductor Manufacturing Co Ltd | 0 | — |
| ProteanTecs Ltd | 0 | — |
| Sony Semiconductor Solutions Corp | 0 | — |
| Semiconductor Energy Laboratory Co Ltd | 0 | — |
| Renesas Electronics Corp | 0 | — |
| Ibiden Co Ltd | 0 | — |
Where to take this next
The dataset points to specific next steps depending on whether the goal is freedom-to-operate, whitespace filing, or competitive tracking.
Check freedom-to-operate against the mid-tier
With filings spread gradually from fifth to tenth place rather than concentrated at the very top, a freedom-to-operate review needs to cover the mid-tier assignees, not just the leader.
Run an FTO scan in EurekaDraft around the thinner subclasses
H10P, H10W, C23C and G11C carry meaningfully lower record shares than H01L and G01R, suggesting claim space still open around packaging, wide-bandgap, and coating-dependent stress mechanisms.
Explore whitespace claims in EurekaTrack co-filing relationships
The ten identified co-assignee pairs cluster around a small set of repeat partners, worth monitoring for signs of expanding joint development programs.
Monitor assignee activity in EurekaCommon questions about this landscape
In this landscape, it is a filing that addresses a specific wearout mechanism — bias temperature instability, hot carrier injection, or time dependent dielectric breakdown — together with a quantitative acceleration or lifetime-extrapolation method such as activation energy, voltage acceleration, or electromigration lifetime modeling. A patent describing only a test structure without an extrapolation method, or only a modeling technique without a named mechanism, generally falls outside this specific intersection. The scope was built to capture the overlap between the physical failure mode and the method used to project device life from stress data, which is narrower than general semiconductor reliability.
The ranking covers 100 companies across 785 records, with the leading assignee holding 90 records and a steep but not exclusionary drop to 34 records at fifth place. Combined, the top 5 assignees account for 37.5% of all 785 records in scope, and the top 10 reach 51.8%. That leaves close to half of all filings coming from outside the ranked leaders, spread across a long tail of smaller and newer entrants.
Filing peaked in 2017 at 58 records and has since settled onto a lower plateau. Looking at the most recent span that can be treated as complete, 2021 to 2024, filings moved from 20 to 18 records, a -10% change — a mild pullback rather than a steep decline. Because publication typically lags filing by roughly 18 months, the 2025 and 2026 figures in any dataset will understate actual filing activity and should not be read as evidence the field is cooling.
This IBM filing claims a test structure and bridge-balance test circuit that simultaneously characterizes electromigration or stress migration failures and time dependent dielectric breakdown failures from a shared via-chain and digitized wire structure. It is significant as prior art for anyone designing a combined test vehicle for these two failure modes in adjacent wiring levels, but it does not cover BTI or hot carrier injection test structures, nor does it claim any specific extrapolation or acceleration-factor method. Teams working on single-mechanism test structures or on the modeling side of lifetime extrapolation are working outside its specific claim scope.
Classification density is noticeably thinner in H10P, H10W, C23C and G11C than in the core H01L and G01R clusters, pointing to packaging-level electromigration structures, wide-bandgap device stress migration, coating-dependent dielectric breakdown acceleration, and memory-array voltage acceleration modeling as comparatively open branches. These are not untouched areas, but they carry a smaller share of the 785 records than the dominant device-structure and measurement classes. A first claim in these branches would likely need to tie a specific material or device architecture to an extrapolation method to differentiate from the denser core clusters.
Research Reliability Physics and Wearout Mechanisms in depth with Eureka
Go past this page: query the whole reliability physics and wearout mechanisms 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.