Fastest
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 →https://www.patsnap.com/resources/blog/rd-blog/computational-and-near-data-storage-patent-landscape/ · Patsnap · data cut-off 2026-07-31 · downloaded from the live page
Patent Landscape · Computational & Near-Data StorageComputational and near-data storage patents: who holds the claim space and where it is still open
- Concentrated at the top. the top 5 assignees combined hold 77.6% of all 98 records in scope, and the top 10 reach 95.9% — a long tail beyond that is thin.
- Filing has cooled from its peak. 2022 was the high point at 23 records; the last complete comparison period, 2021 to 2024, shows a 29% decline (14 to 10).
- Claims cluster in one IPC subclass. G06F covers 89.8% of the 98 records, with H04L (22.4%) and G06N (11.2%) a distant second and third.
98
Published Records
78%
Top-5 Share of All Records
-29%
Filing Growth 2021→2024
US
Leading Jurisdiction
Filing growth compares 2021 (14 records) with 2024 (10) — 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 98 records in scope (CR5), not by the ranked leaders only.
Published byPatsnap Research··7 min readSourced from Patsnap Eureka
Overview
What this dataset covers
This landscape tracks 98 published records matching computational storage, in-storage processing and storage device offload concepts, combined with host interface, programming model, filter pushdown, power budget, data movement reduction and command interface language. Coverage runs from 2015 through the 2026-07-31 data cut-off, with the earliest tracked filing activity in 2017.
Filing activity peaked in 2022 and has since pulled back through the last complete comparison window. Because publication typically lags filing by around eighteen months, the most recent one to two years in any trend line will read lower than the true filing rate and should not be read as a slowdown on their own.
Filing activity, 2017–2026 (2026 partial)
Source: Patsnap Eureka. Assignee ranking and totals. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
The numbers
Filing trend and technology composition
Two views of the same 98 records: how filing activity has moved year over year, and which IPC subclasses the claims actually sit in.
Filing trend
Activity rose from 2 records in 2017 to a peak of 23 in 2022, then declined through the last complete comparison period: 14 records in 2021 versus 10 in 2024, a 29% drop. 2025 and 2026 figures are still filling in given the typical 18-month publication lag.
IPC composition
G06F (electric digital data processing) appears in 89.8% of the 98 records, confirming that most claims are framed as data-processing architecture rather than pure memory or power engineering. H04L (digital transmission, 22.4%) and G06N (AI computing, 11.2%) show meaningful but secondary overlap; G11C (memories) and H02J (power supply) sit at just 2.0% each, marking them as thin coverage areas relative to the core.
Shares are the percentage of the 98 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Source: Patsnap Eureka. Filing trend and technology composition. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.
Go deeper on Computational and Near Data Storage with Eureka
This page is one run against one query. Ask Eureka your own question about computational and near data storage and every answer comes back with the patent numbers behind it.
Try EurekaKey patents
Most-cited records in the corpus
Representative filing
Computational storage for distributed computing (US20180253423A1)
INTERNATIONAL BUSINESS MACHINES CORPORATION
A computational storage server receives input from multiple clients and aggregates it before executing computation functions on the server side. The architecture sets up separate memory areas: one for incoming multi-client input, one for the computation functions themselves, and a client-specific area for metadata tied to each client's requests.Filed by International Business Machines Corporation, published 2018-09-06.
Highest-citation records
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20210144517A1 | Multi-entity resource, security, and service management in edge computing deployments | 840 |
| 2 | WO2020226979A2 | Multi-entity resource, security, and service management in edge computing deployments | 101 |
| 3 | US20200301898A1 | Systems and methods for accelerating data operations by utilizing dataflow subgraph templates | 88 |
| 4 | US20220124009A1 | Intent-based orchestration in heterogenous compute platforms | 52 |
| 5 | US20140137130A1 | Method and apparatus for efficient execution of concurrent processes on a multithreaded message passing system | 52 |
| 6 | US20220121455A1 | Intent-based cluster administration | 36 |
| 7 | US20140136555A1 | Appliance for accelerating graph database management and analytics systems | 34 |
| 8 | US20140137129A1 | Method and apparatus for efficient execution of concurrent processes on a multithreaded message passing system | 31 |
| 9 | US20220113790A1 | Intent-driven power management | 27 |
| 10 | US20140136553A1 | Appliance for accelerating graph database management and analytics systems | 22 |
Citation counts inside a searched corpus favour older records that have had more time to accumulate references — treat this as a signal of influence, not of current technical importance.
Each row carries its publication number; clicking a row searches Eureka by that number.
Source: Patsnap Eureka. Citation counts and representative records. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Run it yourself
Put your own technology through the same analysis
Where to run it
For builders
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 →Insights
What the data means for filing decisions
Three findings that shape where a new filing has room to stand, and where it will compete with dense prior art.
Concentration
77.6%
of 98 records held by top 5 assignees
The field is a leader-plus-tail structure
With the top 5 assignees combined accounting for 77.6% of all 98 records, and the top 10 reaching 95.9%, this is not a fragmented field. The leader alone holds 30 records against a fifth-place figure of 8 and a tenth-place figure of just 2, so most of the remaining ranked companies are single-digit filers.
17 companies make up the entire ranked list returned by the data endpoint.
Momentum
-29%
2021→2024 filing change
Peak activity has passed, for now
Filing rose steadily to a peak of 23 records in 2022, then eased back: the 2021-to-2024 comparison shows a 29% decline, from 14 to 10 records. Because publication lags filing by roughly 18 months, 2025 and 2026 counts are still incomplete and should not be read as confirmation of a continued decline.
2024 is the most recent year treated as a complete filing year.
Technology mix
89.8%
of records in G06F
Claims sit overwhelmingly in data processing architecture
G06F covers 89.8% of the 98 records in scope, dwarfing every other subclass. H04L (22.4%) and G06N (11.2%) show that transmission and AI-adjacent framing are common secondary angles, but subclasses like G11C and H02J sit at just 2.0% each — narrow, memory- and power-specific claim territory that most filers are not touching directly.
Shares sum above 100% because records can carry multiple IPC classes.
Eureka AI Agent
Looking for what nobody has claimed yet?
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to computational and near data storage, with the prior art for and against each one.
Source: Patsnap Eureka. Co-assignee relationships and derived observations. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Players
Who is filing, and where momentum has shifted
The ranked leaders span large device manufacturers, systems vendors and a smaller set of specialist and consumer-platform entrants further down the table.
Leader
30
records
One assignee sets the pace by a wide margin
The top-ranked assignee holds 30 records, roughly four times the fifth-place count of 8. That gap is the clearest sign that a handful of large device and platform companies have built the deepest claim positions in computational storage architecture.
Ranking is based on patent family counts.
Mid-table
8
records at 5th place
A cluster of established players holds the middle
Between 5th and 10th place, filing counts fall from 8 down to 2, covering major storage device, processor and cloud infrastructure vendors. This band is where competitive contests over specific claim language are most likely, since positions are close enough to overlap.
Top 10 combined reach 95.9% of all 98 records.
Momentum
-100%
YoY for several leaders
Recent-year activity has dropped off across the board
Several of the most active historical filers, including major storage and processor manufacturers, show zero records in the latest year with year-over-year changes of -100%, and even the most active recent filer shows a 67% YoY decline. Given the roughly 18-month publication lag, this reads as a data-visibility gap as much as a real pullback.
Momentum figures compare the latest tracked year to the prior year.
Recent-year filing momentum by assignee
| Assignee | Recent year | YoY |
|---|---|---|
| NVIDIA Corporation | 1 | -67% |
| Samsung Electronics Co., Ltd. | 0 | -100% |
| Intel Corporation | 0 | -100% |
| Kioxia Corporation | 0 | — |
| SK Hynix NAND Product Solutions Corp. | 0 | -100% |
| International Business Machines Corporation | 0 | — |
| METSCH THIJS | 0 | — |
| HOBAN ADRIAN | 0 | — |
Source: Patsnap Eureka. Assignee-level momentum. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
What's next
Where to take this analysis
The dataset points to a concentrated core and several thinner branches. The next steps depend on whether the goal is freedom-to-operate or finding open claim space.
Map claims against the leader's portfolio
With one assignee holding 30 of 98 records, any new filing in the core G06F space should be checked against that portfolio specifically before drafting claims.
Explore assignee portfolios in EurekaTest the thinner IPC branches
G11C and H02J sit at 2.0% of records each — worth a closer look to see whether that reflects genuine white space or just a mismatch with the search terms used here.
Run a focused IPC query in EurekaWatch the 2025–2026 publication window
Recent-year momentum figures look sharply negative, but the 18-month publication lag means the true 2025 filing picture is not yet visible in this data.
Track new filings in EurekaSource: Patsnap Eureka. Forward-looking reading of the same dataset. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
FAQ
Common questions about computational storage patents
In this dataset, a computational storage patent is one whose title, abstract or claims combine a core concept such as computational storage, in-storage processing or storage device offload with a functional element like a host device interface, programming model, filter pushdown, power budget management, data movement reduction or command interface. This combination is what distinguishes computational storage filings from ordinary storage-device or general-purpose processor patents. The 98 records in this landscape were identified using exactly this search logic across the 2015 to 2026 window.
The ranked list contains 17 companies, and it is heavily front-loaded: the top 5 combined account for 77.6% of all 98 records, and the top 10 reach 95.9%. The single leading assignee holds 30 records on its own, compared with 8 at fifth place and just 2 at tenth. Large storage device manufacturers, processor vendors and systems companies dominate this list, with a longer tail of smaller and single-filing entrants below it.
Filing activity rose to a peak of 23 records in 2022, then declined through the most recent complete comparison period, falling 29% from 14 records in 2021 to 10 in 2024. Because patent publication typically lags filing by around 18 months, the 2025 and 2026 figures in any dataset are understated and should not yet be read as a genuine slowdown. A clearer read on the current trajectory will only be available once those years finish publishing.
Relative to the dominant G06F (electric digital data processing) subclass, which appears in 89.8% of the 98 records, coverage is much thinner in G11C (static and digital memories) and H02J (power supply and grid systems), each at just 2.0% of records. G06K (data recognition) and H04W (wireless networks) are also comparatively light, at 3.1% and 4.1% respectively. These are candidate areas for narrower, more specific claims rather than broad architecture claims.
US20180253423A1, filed by International Business Machines Corporation and published 2018-09-06, describes a computational storage server that aggregates input from multiple clients before executing computation functions on that input. Its architecture sets up distinct memory areas: one for multi-client input, one for the computation functions themselves, and a client-specific area for request metadata. Anyone building a multi-tenant in-storage compute system that aggregates client inputs before processing should check this filing's claim scope directly rather than relying on the abstract alone.
Answers are grounded in the same dataset. Derived from a Patsnap search on Computational and Near Data Storage covering 2015–2026, data cut-off 2026-07-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Research Computational and Near Data Storage in depth with Eureka
Go past this page: query the whole computational and near data storage 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.