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 →Filing growth compares 2021 (210 records) with 2024 (124) — 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 2,429 records in scope (CR5), not by the ranked leaders only.
Resistive random access memory patenting spans device structure, switching-layer materials, forming-voltage control and selector-device integration. The search behind this page combines core RRAM and filamentary-switching terminology with the operational problems that decide whether a device design is manufacturable: forming voltage, switching variability, retention at temperature, oxygen vacancy migration, selector devices and endurance limits. That combination narrows the corpus to 2,429 records where the claims engage with the engineering trade-offs, not just the general memory-cell concept.
Publication lags filing by roughly 18 months, so the most recent one or two years in any trend understate real filing activity. Read the 2024 figures as the last reasonably complete year, and treat 2025-2026 as still filling in rather than as evidence of a slowdown.
Two views of the same 2,429 records: how filing has moved year over year, and which IPC subclasses the claims fall into.
Filing rose from 190 records in 2017 to a peak of 210 in 2021, then declined to 124 by 2024 — a 41% drop over that three-year span. 2025 and 2026 figures are still incomplete due to publication lag and should not be read as a continued fall.
H01L (semiconductor devices) appears in 64.7% of records, followed by G11C (static and digital memories) at 41.1% and H10N (other electric solid-state devices) at 29.5%. Because records can carry multiple IPC classes, these shares add to more than 100% of the 2,429 records in scope; they describe where claim density sits, not a partition of the corpus.
Shares are the percentage of the 2,429 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
This page is one run against one query. Ask Eureka your own question about resistive random access memory and every answer comes back with the patent numbers behind it.
Try EurekaA resistive random access memory (RRAM) device with a bottom electrode via, a tapered bottom electrode with a base portion and tip portion, a top electrode vertically aligned above it, and a switching layer disposed between the electrodes.Filed by Taiwan Semiconductor Manufacturing Company, Ltd., published 2024-05-16.


| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20140322862A1 | Method of making a resistive random access memory device with metal-doped resistive switching layer | 434 |
| 2 | US20140322885A1 | Method of making a resistive random access memory device | 427 |
| 3 | US9142764B1 | Methods of forming embedded resistors for resistive random access memory cells | 349 |
| 4 | US20080106926A1 | Non-volatile resistance switching memories and methods of making same | 285 |
| 5 | US20080106925A1 | Correlated electron memory | 284 |
| 6 | US7778063B2 | Non-volatile resistance switching memories and methods of making same | 258 |
| 7 | US20100243983A1 | Controlled localized defect paths for resistive memories | 208 |
| 8 | US20140264228A1 | Fin selector with gated rram | 198 |
| 9 | US7639523B2 | Stabilized resistive switching memory | 193 |
| 10 | US9570683B1 | Three-dimensional two-terminal memory with enhanced electric field and segmented interconnects | 189 |
Citation counts inside a searched corpus favour older filings that have had more time to accumulate citations; treat them as a signal of influence on the field, not a ranking of current technical importance.
Each row carries its publication number; clicking a row searches Eureka by that number.
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 →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 →Three read-outs from the dataset that matter more for strategy than the raw counts alone.
Just five assignees account for just over half of all records in scope, and the top ten extend that to 72.4%. A newcomer is not entering an open field; it is entering one where a handful of players have built dense, overlapping positions around structure and switching-layer composition.
Filing peaked in 2021 at 210 records and had fallen to 124 by 2024, a 41% decline over three years. Several of the most active historical filers show sharply reduced or zero filings in the latest tracked year, which is consistent with either a maturing core design space or a shift of effort toward adjacent claim types not fully captured by this search string.
H01L semiconductor-device claims cover 64.7% of records while G06F electric digital data processing claims sit at only 6.5% and H03K pulse-technique and logic-circuit claims at 3.4%. The device layer is heavily claimed; the compute and logic integration layer around RRAM is comparatively open.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to resistive random access memory, with the prior art for and against each one.
The ranking covers the whole set of 100 companies the data endpoint returns, not a curated top list. A small group holds most of the volume; most of the ranked group holds far less.
The leading assignee's 365 records compare to 151 at fifth place and 60 at tenth — a steep drop-off rather than a gradual one, meaning the leader's position is not simply a product of broad-based industry filing.
With the top ten already accounting for nearly three-quarters of all records, the remaining 90 ranked assignees divide the rest — a long tail of narrower, more specific filing positions rather than broad platform claims.
Several assignees with substantial historical filing show steep year-over-year declines or zero filings in the latest tracked year. Given the roughly 18-month publication lag, this is as likely to reflect filings still working through the pipeline as an actual retreat from the space.
| Assignee | Recent year | YoY |
|---|---|---|
| Micron Technology, Inc. | 4 | -43% |
| Taiwan Semiconductor Manufacturing Co., Ltd. (TSMC) | 1 | -94% |
| International Business Machines Corporation (IBM) | 0 | -100% |
| Intermolecular, Inc. | 0 | — |
| Crossbar, Inc. | 0 | -100% |
| Winbond Electronics Corporation | 0 | — |
| Intel Corporation | 0 | — |
| Toshiba Corporation | 0 | — |
The dataset points to specific next steps depending on whether the goal is freedom-to-operate, portfolio strategy, or identifying open claim space.
With over half of all records held by five assignees, a freedom-to-operate review for any RRAM device or integration scheme should start with the leading filers' claim scope before looking at the long tail.
Explore assignee claims in EurekaG06F and H03K shares are low relative to the core H01L device layer, suggesting compute-in-memory and logic-integration claims around RRAM are comparatively open.
Run a white-space search in EurekaBecause 2025-2026 records are still incomplete, re-check momentum by assignee again once another 12-18 months of publications land, rather than treating the current dip as final.
Track filing trends in EurekaThe leading assignee in this dataset holds 365 records, well ahead of the fifth-ranked filer at 151 and the tenth at 60. The top five assignees combined hold 51.0% of all 2,429 records in scope, and the top ten hold 72.4%. This means filing is heavily concentrated among a small group rather than spread evenly across the 100 companies in the ranked list, so any competitive review should start with the leaders and treat the remaining ranked assignees as a long tail of narrower positions.
Filing peaked in 2021 at 210 records and declined to 124 by 2024, a 41% drop over three years — the most reliable read available because 2024 is the last year the dataset treats as complete. Figures for 2025 and 2026 are lower still, but that reflects the roughly 18-month lag between filing and publication rather than a real collapse in activity. Anyone tracking this trend should wait for another filing cycle to complete before treating the recent years as a genuine slowdown.
H01L (semiconductor devices) appears in 64.7% of the 2,429 records in scope, making it the dominant classification, followed by G11C (static and digital memories) at 41.1% and H10N (other electric solid-state devices) at 29.5%. Because a single record can carry several IPC classes, these percentages add up to more than 100% of records — they show where claim density concentrates, not a clean split of the field. Lower-density classes like G06F (6.5%) and H03K (3.4%) mark areas with comparatively less claim coverage.
The clearest signal is the gap between core device claims and system-level claims: H01L device-structure claims cover 64.7% of records while G06F data-processing claims sit at only 6.5% and H03K logic-circuit claims at 3.4%. This points to under-claimed territory around compute-in-memory circuits, selector-device co-integration, and endurance or retention compensation schemes built around RRAM cells rather than the cell structure itself. A first claim in these areas would need to tie a specific circuit-level function to the RRAM switching behaviour rather than claiming the device structure alone.
US20240164225A1, filed by Taiwan Semiconductor Manufacturing Company and published 2024-05-16, claims an RRAM device built around a tapered bottom electrode with a distinct base and tip portion, a vertically aligned top electrode, and a switching layer positioned between them. It blocks designs that use this specific tapered bottom-electrode geometry and electrode alignment, but it does not cover RRAM devices built on different electrode shapes, different switching-layer materials, or different integration schemes such as selector-device pairing. Anyone designing a bottom-electrode structure should compare their geometry against this claim closely rather than assuming all tapered-electrode approaches are equally exposed.
Go past this page: query the whole resistive random access memory 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.
Help us improve
Something wrong, out of date, or missing on this page? Tell us and it goes straight to the team that maintains it.
This page