GaN HEMT CMP Patents: Who Leads, Where the Gaps Are 2026
- Filing has cooled since its 2019 peak. 58 families that year against 23 in 2017 and a flat-to-declining midpoint of 19 in 2022 — this is not a growth curve, it's a settled claim space.
- US filings dominate the venue mix. 218 of the tracked records were filed at the USPTO, more than the EPO, WIPO, Israel, Germany and Taiwan combined.
- No single assignee shows recent-year momentum. Every major name tracked — from large-cap chipmakers to specialist GaN foundries — logged zero filings in the latest year, consistent with publication lag rather than an exited field.
Filing growth compares 2021 (47 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 338 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This landscape tracks patent families at the intersection of gallium nitride HEMT device fabrication and chemical mechanical planarization (CMP) — the polishing step that flattens dielectric and metal layers between process stages. The search spans GaN HEMT, GaN transistor and GaN surface planarization language paired against CMP process terminology, across 338 published patent families filed between 2015 and mid-2026.
Because publication typically lags filing by around 18 months, the most recent one or two years in any trend chart will always look thinner than they eventually turn out to be — that softens, but does not erase, the decline visible from the 2019 peak.
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 338 families: how filing volume has moved year over year, and which IPC subclasses carry the claim density.
A peak already behind it
Filings rose from 23 in 2017 to a peak of 58 in 2019, then eased back toward a midpoint of 19 by 2022. The curve reads as a technology that had its land-grab phase and has since settled into maintenance-level filing, rather than one still accelerating.
Concentrated in core semiconductor-device classes
H01L (semiconductor devices) accounts for 297 of the 338 records, with H10D close behind at 142 — the modern subclass split-off for general semiconductor device structures. Smaller pockets sit in H10W and H10P (process/packaging-adjacent classes), H10N (other solid-state devices), H01S (lasers), B32B (layered products) and H02M (power conversion), showing CMP-for-GaN claims reaching into optoelectronic and power-conversion applications but not dominating them.
Shares are the percentage of the 338 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Gallium Nitride HEMT Chemical Mechanical Planarization with Eureka
This page is one run against one query. Ask Eureka your own question about gallium nitride hemt chemical mechanical planarization and every answer comes back with the patent numbers behind it.
Try EurekaA representative recent filing
BEOL integration solution based on direct CMP to improve intermetal dielectric layer
A process that helps ensure uniform height of conductive structures formed among intermetal dielectric layers of a wafer. When a metal layer is deposited on a first intermetal dielectric layer, a sealing layer is formed on the metal layer either before or after the metal layer is patterned to form metal interconnect structures. A first interlevel dielectric sub-layer is then formed on the sealing layer. A chemical mechanical planarization (CMP) process is then performed on the first interlevel dielectric sub-layer using the sealing layer as an etch stop. A second interlevel dielectric sub-layer is then formed on the first interlevel dielectric sub-layer.Filed by STMicroelectronics International N.V., published 2025-05-22 — one of the most recent entrants in this dataset, illustrating the back-end-of-line (BEOL) integration angle on CMP rather than front-end device planarization.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US9653642B1 | Manufacturable RGB display based on thin film gallium and nitrogen containing light emitting diodes | 168 |
| 2 | US6862127B1 | High performance micromirror arrays and methods of manufacturing the same | 160 |
| 3 | US9666677B1 | Manufacturable thin film gallium and nitrogen containing devices | 137 |
| 4 | US10002928B1 | Manufacturable RGB display based on thin film gallium and nitrogen containing light emitting diodes | 119 |
| 5 | US8304271B2 | Integrated circuit having a bulk acoustic wave device and a transistor | 116 |
| 6 | US20060284247A1 | Novel method for integrating silicon CMOS and AlGaN/GaN wideband amplifiers on engineered substrates | 100 |
| 7 | US20070018198A1 | High electron mobility electronic device structures comprising native substrates and methods for making the s… | 88 |
| 8 | US6903860B2 | Vacuum packaged micromirror arrays and methods of manufacturing the same | 79 |
| 9 | US20170309676A1 | Engineered Substrate Including Light Emitting Diode and Power Circuitry | 78 |
| 10 | US20110248283A1 | Via structure of a semiconductor device and method for fabricating the same | 69 |
Citation counts favour older, more-searched records inside this corpus — read them as a signal of influence on the field, not of current filing priority.
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 numbers mean for a filing decision
Three patterns stand out once family counts, venues and IPC composition are read together.
The claim space is settled, not open
A run-up to 58 families in 2019 followed by a decline to 19 by 2022 signals that the foundational CMP-for-GaN-HEMT process moves were staked out early. New entrants now compete for narrower process variants rather than broad method claims.
US filing is the default venue
The USPTO share (218) dwarfs EPO (44), WIPO/PCT (42) and the smaller national offices. Freedom-to-operate work that skips a US search misses the bulk of the enforceable art in this field.
Device-structure classes carry the density
H01L and its H10D successor account for the large majority of records, meaning most claims are anchored to semiconductor device structure rather than to standalone polishing chemistry or equipment — a clue for where narrower, process-only claims might still clear.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to gallium nitride hemt chemical mechanical planarization, with the prior art for and against each one.
| Assignee | Co-assignee | Shared families |
|---|---|---|
| ZUNIGA MARCO A | YANG LU | 10 |
| ZUNIGA MARCO A | RUAN JUN | 10 |
| ZUNIGA MARCO A | PRASAD JAYASIMHA | 10 |
| ZUNIGA MARCO A | PAUL AMIT | 10 |
| YANG LU | RUAN JUN | 10 |
| YANG LU | PRASAD JAYASIMHA | 10 |
| YANG LU | PAUL AMIT | 10 |
| RUAN JUN | PRASAD JAYASIMHA | 10 |
Only 10 co-assignee pairs were found across the dataset, with the strongest links (each appearing 10 times) centred on a single small cluster of individual inventors — a sign that most filings here are single-assignee efforts rather than joint development.
Who holds the ground
Recent-year filing counts across the tracked assignees — from diversified chipmakers to specialist GaN foundries — all read zero in the latest year, which is expected given publication lag but still means momentum has to be judged on the trend line, not the newest data point.
No visible front-runner in the newest data
Every named assignee tracked for recent-year momentum, including large diversified semiconductor firms and dedicated GaN specialists, shows zero filings in the most recent year and one shows a -100% year-over-year drop. This is consistent with the 18-month publication lag rather than an actual pullback from the technology.
Filing here is mostly a solo activity
With only 10 co-assignee pairs identified across 338 families, and the strongest links tied to a small inventor cluster rather than cross-company alliances, joint development deals are the exception in this landscape, not the norm.
Watch US prosecution first
Given the US filing skew, competitive monitoring and freedom-to-operate diligence should prioritise USPTO file histories and continuations before extending to EPO or PCT-only families.
| Assignee | Recent year | YoY |
|---|---|---|
| Intel Corporation | 0 | — |
| Taiwan Semiconductor Manufacturing Company (TSMC) | 0 | -100% |
| Vanchip Semiconductor Co., Ltd. | 0 | — |
| Raytheon Company | 0 | — |
| Innoscience (Suzhou) Technology Co., Ltd. | 0 | — |
| Qualcomm Incorporated | 0 | — |
| GlobalFoundries U.S. Inc. | 0 | — |
| ZUNIGA MARCO A | 0 | — |
Where to take this research
The trend and composition data point to a mature but not closed claim landscape. Two directions are worth pursuing depending on your role.
Run a freedom-to-operate check on the under-claimed branches
Slurry chemistry, post-CMP passivation and endpoint detection on non-planar topography show thinner IPC density than the core device-structure classes — worth confirming before committing engineering resources.
Explore in EurekaTrack US prosecution histories, not just grants
With 218 of 338 families filed at the USPTO, continuation and file-wrapper monitoring there will surface competitive moves earlier than watching grant announcements alone.
Set up monitoring in EurekaCommon questions on GaN HEMT CMP patents
Chemical mechanical planarization (CMP) flattens dielectric and metal layers deposited during GaN HEMT fabrication so that subsequent lithography and deposition steps land on a uniform surface. In this dataset, the bulk of claims sit in semiconductor-device IPC classes (H01L, H10D) rather than standalone polishing-equipment classes, meaning most patented CMP work here is tied directly to a specific device structure step, such as intermetal dielectric layers or gate stack formation, rather than a general-purpose polishing method. That matters for drafting: a claim scoped only to the polishing chemistry or tool, independent of the GaN device context, is less common and may face less prior art in this specific dataset.
Yes, based on the tracked filing trend. Filings rose from 23 in 2017 to a peak of 58 in 2019, then declined to a midpoint of 19 by 2022, and the trend has stayed flat to declining since. This pattern is typical of a technology area that had an initial land-grab period followed by a shift toward narrower, incremental filings rather than foundational method claims. Readers should still account for publication lag of roughly 18 months, which understates the most recent one to two years of any trend.
Start with the USPTO. Of the 338 tracked patent families, 218 were filed there, far ahead of the European Patent Office (44) and WIPO/PCT filings (42), with smaller counts at the Israeli, German and Taiwanese offices. A freedom-to-operate search limited to non-US offices would miss the majority of the enforceable art in this specific technology area.
The recent-year momentum data shows every tracked assignee, including large diversified chipmakers and specialist GaN foundries, at zero filings in the latest year, with at least one showing a -100% year-over-year change. This is consistent with the roughly 18-month lag between filing and publication rather than a confirmed exit from the technology, so it is safer to read this as an incomplete latest-year picture than as evidence that the field has gone cold.
The IPC composition thins out noticeably outside the core H01L and H10D device-structure classes. Areas such as CMP slurry chemistry specific to AlGaN/GaN interfaces, post-CMP surface passivation for HEMT gate stacks, and CMP endpoint detection tuned to non-planar HEMT topography show far fewer claims relative to the core device classes. These branches are worth a dedicated novelty search before assuming they are unclaimed, since low density in a landscape search does not guarantee an empty prior-art field, only a thinner one.
Research Gallium Nitride HEMT Chemical Mechanical Planarization in depth with Eureka
Go past this page: query the whole gallium nitride hemt chemical mechanical planarization 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.