Neuromorphic Computing Patents: Top Companies & Filing Trends 2026
- 77.3% of all 5,042 records sit with just five assignees, and 85.8% with ten — this field is unusually concentrated for a computing sub-domain still this young.
- Filings peaked in 2021 at 924 then fell to 497 by 2024, a documented -46% swing that predates the usual 18-month publication lag on the most recent years.
- G06F and G06N cover the core, but H04L, G06T and G05B each carry a meaningful share — signalling that neuromorphic claims increasingly reach into networking, vision and control-system integration.
Filing growth compares 2021 (924 records) with 2024 (497) — 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 5,042 records in scope (CR5), not by the ranked leaders only.
What the neuromorphic computing patent record actually shows
Neuromorphic computing patenting sits at the intersection of spiking neural network architectures and conventional compute-acceleration concerns — cache coherence, memory access, task parallelism and workload acceleration all appear as required companions in the search definition behind this dataset. That pairing matters: the records here are not abstract neuroscience-inspired designs, they are filings that tie spiking or neuromorphic processing to concrete system-level performance problems. The dataset spans 2015 through the 2026 cut-off and totals 5,042 published records, treated here as patent families for ranking purposes.
The assignee base is short and heavily weighted at the top, the technology composition leans on general digital data processing and AI-model computation classes, and the United States dominates as a receiving office by a wide margin over the WIPO, UK and EPO routes. Each of these patterns has direct implications for where a new filer can still stake a defensible claim.
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 5,042-record corpus: how filing activity has moved year over year, and which IPC subclasses the claims actually sit in.
Filing trend, 2017–2026
Annual filings rose from 71 in 2017 to a peak of 924 in 2021, then declined to 497 by 2024 — a -46% move over that three-year span. 2025 and 2026 figures are still incomplete due to the roughly 18-month lag between filing and publication, so the recent-year dip should not be read as the field cooling.
Technology composition by IPC subclass
G06F (electric digital data processing) appears in 54.9% of records and G06N (AI-model computation) in 38.0%, confirming the core is genuinely compute-and-learning-architecture work. Smaller but non-trivial shares in H04L (16.6%), G06T (12.7%), G06K (8.0%), G05B (8.0%), G06V (7.6%) and G06Q (6.5%) show the technology reaching into networking, vision, recognition, control and commerce applications — since records can carry multiple classes, these shares sum to more than 100%.
Shares are the percentage of the 5,042 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Neuromorphic Computing Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about neuromorphic computing patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative and most-cited filings
System and method for mapping of spiking neural networks on neuromorphic processor
A method for mapping a spiking neural network design onto a configurable neuromorphic processor: defining a resource model describing the processor, defining a network definition file describing the network design, dividing the network's neurons into partitions aligned to layers, selecting a partition, selecting an available plane of the resource model with sufficient neuron capacity, and mapping the selected partition's neurons onto that plane.Filed by Innatera Nanosystems B.V., published 2024-08-29 as WO2024175767A1 — a recent example of hardware-mapping claims for spiking neural network deployment on configurable neuromorphic silicon.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20190339688A1 | Methods and systems for data collection, learning, and streaming of machine signals for analytics and mainten… | 974 |
| 2 | US20200348662A1 | Platform for facilitating development of intelligence in an industrial internet of things system | 732 |
| 3 | US20210157312A1 | Intelligent vibration digital twin systems and methods for industrial environments | 715 |
| 4 | US20180284758A1 | Methods and systems for industrial internet of things data collection for equipment analysis in an upstream o… | 600 |
| 5 | US20200225655A1 | Methods, systems, kits and apparatuses for monitoring and managing industrial settings in an industrial inter… | 563 |
| 6 | US20200103894A1 | Methods and systems for data collection, learning, and streaming of machine signals for computerized maintena… | 514 |
| 7 | US20210342836A1 | Systems and methods for controlling rights related to digital knowledge | 468 |
| 8 | US20220366494A1 | Market orchestration system for facilitating electronic marketplace transactions | 438 |
| 9 | CN112703457A | 用于使用工业物联网进行数据收集、学习和机器信号流传输实现分析和维护的方法和系统 | 412 |
| 10 | US20190033845A1 | Methods and systems for detection in an industrial internet of things data collection environment with freque… | 409 |
Citation counts favour older filings simply because they have had more time to accumulate citations within the searched corpus — treat them as a signal of influence, not of current technical importance.
Patent titles are shown in the language they were filed in, not translated, so that each record stays verifiable against the original filing — a translated title will not match in Eureka or in any national register. 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 the concentration, trend and classification data are read together.
Five assignees hold more than three-quarters of the field
With 3,898 of 5,042 records in the hands of just five companies, freedom-to-operate work in neuromorphic computing has to start with those portfolios specifically, not with a generic prior-art sweep. The next five names add only another 8.5 percentage points, so the drop-off after the leaders is steep.
Activity has cooled from its 2021 peak, but the picture is incomplete
Filings fell from 924 in 2021 to 497 in 2024. Because publication trails filing by roughly 18 months, 2025 and 2026 figures will keep revising upward, so this is a real but partial retreat from a filing surge rather than confirmed evidence the technology is stalling.
Claims are anchored in compute infrastructure, not just AI models
The dominance of G06F (54.9%) over the AI-specific G06N (38.0%) shows most filings frame neuromorphic work as a systems and data-processing problem first. That framing shapes where examiners will look for prior art and where claim drafting needs to distinguish from general-purpose accelerator patents.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to neuromorphic computing patent landscape, with the prior art for and against each one.
Who is filing, and where the gate stands
The leader board is short, the momentum is uneven, and several architectural branches remain thin enough to be worth watching.
A single assignee holds roughly a third of the entire corpus
The top-ranked assignee's 1,603 records dwarf the fifth-place total of 151, meaning the leader's portfolio alone shapes much of the prior-art landscape any new filer has to navigate.
Filing drops off fast after the top names
Tenth place holds only 57 records against the leader's 1,603, and the ranking of 100 companies overall shows a sharp long tail rather than a broad, evenly distributed competitive field.
Even leading filers are pulling back year over year
Recent-year momentum shows steep YoY declines across several of the largest assignees, alongside at least one smaller portfolio holder posting a +100% YoY increase off a low base — a reminder that late-stage entrants can still move fast in a concentrated field.
| Assignee | Recent year | YoY |
|---|---|---|
| NVIDIA Corp | 26 | -77% |
| Pure Storage Inc | 20 | -86% |
| Innatera Nanosystems B.V. | 4 | 0% |
| STRONG FORCE TX PORTFOLIO 2018 LLC | 2 | +100% |
| Intel Corp | 1 | -89% |
| Samsung Electronics Co., Ltd. | 1 | -89% |
| Strong Force IoT Portfolio 2016 LLC | 0 | -100% |
| Strong Force VCN Portfolio 2019 LLC | 0 | -100% |
Where to take this from here
The dataset points to three practical next steps for a team deciding where to file or where to watch.
Map freedom-to-operate against the top five portfolios first
Because 77.3% of records sit with five assignees, a targeted review of those five portfolios will surface most of the relevant prior art faster than a broad keyword sweep across the full 100-company ranking.
Explore assignee portfolios in EurekaTrack the under-claimed branches before they fill in
Overlaps with G05B, G06Q and G06V suggest room for claims that combine neuromorphic architectures with control, commerce-data or vision workloads — areas thinner than the G06F/G06N core.
Run a white-space search in EurekaRevisit the 2024-2026 filing trend once publications catch up
The apparent -46% drop from 2021 to 2024 is real, but 2025-2026 figures will keep rising as publications catch up to filings. Re-check the trend in twelve months before drawing conclusions about slowing demand.
Set a trend alert in EurekaCommon questions about the neuromorphic computing patent landscape
It is highly concentrated: the top five assignees hold 3,898 of the 5,042 records in scope, which is 77.3% of the field, and the top ten hold 85.8%. That means the vast majority of prior art relevant to a freedom-to-operate review sits with a small number of companies rather than being spread across many competitors. A new entrant doing clearance work should prioritise those leading portfolios before running a broader search across the full ranked list of 100 companies.
Filings rose sharply to a peak of 924 in 2021, then fell to 497 by 2024, a documented -46% change over that span. However, because patent publication typically lags filing by around 18 months, the 2025 and 2026 figures in any dataset are still incomplete and will revise upward over time. The honest read is that filing activity has cooled from its 2021 high, not that the technology is being abandoned.
The largest classes are G06F (electric digital data processing), covering 54.9% of the 5,042 records, and G06N (computing based on AI models), covering 38.0%. Smaller but meaningful shares appear in H04L (digital information transmission, 16.6%), G06T (image data processing, 12.7%), G06K (data recognition, 8.0%), G05B (control systems, 8.0%), G06V (image/video recognition, 7.6%) and G06Q (business data processing, 6.5%). Because a single record can carry several classes, these percentages add up to more than 100% and should not be summed.
The ranked list covers 100 companies, with the top assignee holding 1,603 records against 151 at fifth place and just 57 at tenth — a steep long tail rather than a broad, even field. Several of the largest filers are showing sharp year-over-year declines in the most recent year, while at least one smaller portfolio holder has grown its filing rate off a low base. This combination of a dominant leader and uneven recent momentum is worth tracking rather than assuming the current leaderboard is static.
The classification data points to thinner claim density where neuromorphic architectures overlap with control systems (G05B), business/commerce data processing (G06Q) and vision recognition (G06V) compared with the dense G06F/G06N core. Cross-node cache coherence for neuromorphic clusters and spiking-network mapping onto reconfigurable silicon planes also appear less saturated based on the available filings. These are not guarantees of patentability, but they are branches where the existing claim density is measurably lower than the core computing classes.
Research Neuromorphic Computing Patent Landscape in depth with Eureka
Go past this page: query the whole neuromorphic computing patent landscape 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.