3D NAND Flash Scaling Patents: Who Leads, Where the Gaps Are 2026
- Filing activity peaked at 11 records in 2023 after years of near-zero output, then dropped back — a single spike rather than a sustained ramp.
- 16 published families total concentrated almost entirely under IPC class H10B, meaning the searchable claim space here is still small and specific.
- United States drew 10 of the filings against 4 in China and 2 in Taiwan, so US filing strategy dominates this corpus more than global manufacturing share would suggest.
What this landscape covers
3D NAND scaling depends on stacking more memory layers without stacking more cost, and the claim activity here tracks three mechanical answers to that problem: string stacking across multiple decks, wafer bonding that separates array and logic fabrication, and CMOS-under-array integration that places control circuitry beneath the memory stack. The search also captures system integration claims spanning staircase contacts, interconnect structures and the mechanical/chemical steps used to planarise and connect them.
The dataset is small and narrow by design: 16 published families sit almost entirely inside IPC subclass H10B, with only light overlap into general semiconductor device claims (H01L) and process steps like grinding, polishing and polishing compositions. That concentration makes this a workable landscape to read claim-by-claim rather than statistically.
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
Publication counts by year and by IPC subclass, drawn from the 16 families matching the search string.
A spike, not a ramp
Filings sat at zero through the middle of the period, then rose to a peak of 11 records in 2023 before falling away again. Because publication lags filing by roughly 18 months, the most recent years understate real filing activity — but the shape here is a burst around one filing cycle, not a steady climb across the window.
Claim space is narrow
All 16 records sit in H10B (memory device manufacture); a handful also touch H01L (semiconductor devices generally) and single records touch grinding, polishing compositions and adjacent H10-series subclasses. There is almost no spread into unrelated fields, which means the white space to look for is inside H10B, not outside it.
Shares are the percentage of the 16 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on 3D NAND Flash Scaling and Integration with Eureka
This page is one run against one query. Ask Eureka your own question about 3d nand flash scaling and integration and every answer comes back with the patent numbers behind it.
Try EurekaA representative claim: staircase contact structures
Memory device (US20250107081A1)
The claim covers a memory device built around an interconnect structure, a staircase structure, a dielectric layer and a stop structure sitting between the interconnect and the staircase/dielectric layers. The stop structure carries an opening that exposes the interconnect, and a first contact runs through the dielectric layer and that opening to reach it.Published 2025-03-27. This is a structural fix for a known failure mode: contacts landing on staircase steps can punch through into the interconnect below if there is no etch-stop layer defining where the contact must land.


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 filing strategy
Three readings of the same 16-family dataset, aimed at where to file next and where not to bother.
One cycle, not a trend
With a midpoint year of 2022 at zero and a single peak of 11 filings in 2023, this looks like one filing wave tied to a specific product or process milestone rather than sustained R&D output. Treat the 2023 spike as the signal to investigate, not as evidence of ongoing momentum.
US-weighted filing
Ten of sixteen records were filed as US applications, with China and Taiwan taking the rest. For a manufacturing process this globally distributed, that skew suggests either a US-centric applicant base in this narrow search or a deliberate US-first filing strategy for defensive coverage.
Claim space is shallow but dense
Every family in this set touches H10B, and the overlap into H01L, grinding and polishing subclasses is minimal. That means competitors are not spreading claims across adjacent process domains — the contest is happening inside a tight structural claim space around staircase, stacking and bonding geometry.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to 3d nand flash scaling and integration, with the prior art for and against each one.
Who is filing, and where the gaps sit
With only 16 families in this corpus, no single assignee shows sustained multi-year output — momentum here is better read filing-by-filing than by cumulative rank.
Flat recent output from the named assignee
The one assignee with recent-year data tracked shows zero filings in the latest year, consistent with the broader post-2023 drop-off across the dataset rather than a company-specific pullback.
US filings dominate the corpus
Given the small sample size, US-heavy filing may reflect which applicants and search terms surface in this dataset rather than a global filing preference across the whole 3D NAND industry.
Everyone is claiming the same subclass
The absence of spread into adjacent IPC codes means differentiation between filers has to be read at the claim level — staircase geometry, stop-layer placement, bonding sequence — not at the subclass level.
| Assignee | Recent year | YoY |
|---|---|---|
| Macronix International Co., Ltd. | 0 | — |
Where to take this analysis
The trend and classification data point to a narrow, structurally-focused claim space with one recent filing spike. These are the natural next steps for a team acting on it.
Chart the staircase and stop-layer claims
With all 16 families sitting inside H10B, the differentiating detail is in structural claim language around staircase contacts, stop layers and interconnect routing — worth a claim chart before filing near this geometry.
Run a claim comparison in EurekaConfirm the 2023 spike against priority dates
Publication lag means the apparent drop after 2023 may partly be unpublished recent filings. Pulling priority dates rather than publication dates will show whether the wave has actually ended.
Check priority filings in EurekaTest the white-space branches directly
Etch-stop placement, bonding alignment tolerance and post-bond polishing chemistry show minimal coverage in this corpus — each is a candidate for a freedom-to-operate check before drafting new claims there.
Search white space in EurekaCommon questions on 3D NAND scaling patents
The core classification is H10B, which covers memory device manufacture and captures essentially all of the structural and process claims around 3D NAND stacking, staircase contacts and bonding. Related claims also appear under H01L for general semiconductor device structure, and under process-specific classes like B24B (grinding and polishing) and C09G (polishing compositions) when the claim covers a fabrication step rather than the device itself. Searching H10B alone will miss claims that are framed as a polishing or bonding process rather than a device structure, so both angles are worth checking.
This dataset shows a peak of 11 published records in 2023 after a flat period with effectively zero filings at the 2022 midpoint. A single-year spike like this usually corresponds to a specific product generation or process node reaching a filing deadline, rather than a sustained increase in R&D investment. Because publication lags actual filing by roughly 18 months, the apparent drop-off after 2023 should be checked against priority dates before concluding that filing activity has genuinely slowed.
CMOS-under-array is an integration approach where the control logic circuitry is built underneath the memory cell array rather than beside it, which saves die area as NAND stacks grow taller. It appears in this search because it is one of the named system-integration techniques used to keep 3D NAND scaling economical as layer counts increase. Claims in this space tend to focus on thermal budget management and interconnect routing between the logic layer and the array above it, both of which show limited coverage in this particular dataset.
The representative filing in this dataset, US20250107081A1 from Macronix International, claims a specific staircase contact structure that uses a dedicated stop layer to prevent contacts from punching through into the interconnect below. This is a narrow structural fix rather than a broad staircase-architecture claim, so it blocks that specific etch-stop arrangement without covering staircase contacts generally. Because the whole corpus sits inside one IPC subclass with limited assignee diversity, a full claim chart across the 16 families is needed to see which staircase approaches remain open.
Based on the IPC spread in this corpus, wafer-bonding alignment tolerance and post-bond polishing chemistry are only lightly claimed compared with the core staircase and stacking structures. That does not mean the space is empty — it means this particular search, limited to string stacking, wafer bonding, CMOS-under-array and system integration terms, surfaces few claims there. A team considering filing in bonding alignment or polishing composition chemistry should run a dedicated search on those terms alone before assuming the space is open, since a narrow search like this one can understate coverage that exists under different classification.
Research 3D NAND Flash Scaling and Integration in depth with Eureka
Go past this page: query the whole 3d nand flash scaling and integration 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.