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 →1,143 patent families cover activation-function hardware; IBM and Intel share the top filer spot at 37 families each.
2,369 published records collapse into 1,143 simple (DOCDB) families; the difference is the same invention filed in more than one country. INTERNATIONAL BUSINESS MACHINE CORPORATION is named on 37 of the 1,143 patent families in scope. One patent family can name several filers, so these figures are not slices of a divided whole and do not add up. In force = 1,296 of 2,369 records still granted and maintained at the data cut-off; the rest are pending, withdrawn, refused or lapsed.
This landscape is built from a search string tying activation function and nonlinear function terms to neural network or artificial neuron language, combined with hardware, circuit or accelerator terms. The result set holds 2,369 published records, which de-duplicate down to 1,143 simple patent families once filings across multiple countries for the same invention are collapsed. The search carries no date limit, so the set spans whatever filing years the matching records happen to cover.
Of the 2,369 published records, 1,296 (55%) were granted and still in force at retrieval. The rest are pending, withdrawn, refused, lapsed or not determinable, and that breakdown is not available here.
Ordered by volume only, not by importance or claim strength. Counted in simple (DOCDB) patent families, so one invention filed in several countries counts once. Names come from Patsnap's standardised assignee field as recorded on each document; subsidiaries and group companies are not merged. The icon before each name is read from the name itself (words such as University, Institute or Ltd); the data carries no organisation-type field, so treat it as a reading aid rather than a classification. 2 rows in this table start with the same name (AMAZON TECH INC, AMAZON COM SERVICES LLC). They are left as separate rows, so the largest-filer figure above is per legal entity, not per group. Rows that look related still cannot be added together: one family can name several of them. Filings held by named individuals rather than organisations are removed where they can be identified (1 here); that identification is not exhaustive, so an individually held filing can still appear below. Either way those filings stay inside the totals above. This table shows the top 30 of the 99 filers returned for this search; the data endpoint returns at most 100 entities, so even that 99 is not the complete set of filers. A single family can name more than one filer, so these counts overlap and do not add up to the total.
Pick a task. Every answer cites the patents behind it.
Two views of the same 1,143-family pool: how published records have moved by filing year, and how the IPC subclasses line up against the family total.
Published records climbed from 4 in 2009 to a peak of 346 in 2021, with 2018's 196 rising to 2021's 346, a 77% increase over that three-year stretch. Records filed from 2022 are still publishing, so the most recent bars understate eventual volume and should not be read as a slowdown.
Counted in published records by earliest filing year, not de-duplicated into families. A filing is counted in the year it was filed, but it only becomes visible after it publishes: roughly 18 months for most routes, and up to 30 months where a PCT application enters national phase. 2022 onwards are therefore still filling in and are shaded; no growth figure on this page uses them. Because this series counts documents rather than inventions, a rise can mean more inventions or the same inventions filed in more countries; this dataset cannot separate the two. This chart covers 2009-2026 (2,251 records); the 2,369 total quoted above includes earlier filings outside the window.
G06N (computing based on AI models) is named on 900 of the 1,143 families (78.7%), and G06F (electric digital data processing) on 396 (34.6%). Because one family can carry several IPC classes, these shares add to more than 100% and are not a partition of the set.
Shares are the percentage of the 1,143 patent families in scope. One patent family 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 neural network activation function hardware implementation patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaThe abstract describes an artificial neural network built around a compute-in-memory accelerator, with a hidden layer of artificial neurons trained using a bitwise modified rectified linear unit activation function. That function generates an output proportional to the input below a critical threshold, and an output independent of the input above it.US12373686B2, filed by Samsung Electronics Co., Ltd., published 2025-07-29. This is simply the first record returned under the database's default order, not a ranking by citations, breadth or date.


Selected as the first record the search returns under the database default order, then shown in full. It is not ranked by citations, claim breadth or commercial importance, and it is not a judgement about which filing in this space matters most.
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 figures worth sitting with before reading the filer table or the IPC chart.
International Business Machine Corporation and Intel Corp are each named on 37 of the 1,143 simple patent families (3.2% each). Filer rows are not rolled up to corporate group and one family can name several filers, so these rows cannot be summed into a combined share.
9 of the 99 listed filers hold 2 or fewer simple patent families each. The list itself covers organisations, filings held by named individuals are excluded from it where they can be identified but remain inside every total on this page.
900 of the 1,143 families (78.7%) carry the G06N subclass, and 396 (34.6%) carry G06F. Because families can carry more than one IPC class, these two shares overlap rather than add, and together they still do not account for every family.
1,296 of 2,369 published records (55%) were granted and still in force at the 2026-10-27 retrieval date. This is a snapshot of legal status, not a measure of how many filings succeeded or failed, and it will keep shifting as records lapse, get challenged or grant.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to neural network activation function hardware implementation patent landscape, with the prior art for and against each one.
This page reports what one search string returned. Narrower questions need a narrower search.
A family count says nothing about what a filer's claims actually cover. Pull the granted claim set for any filing that matters before drawing conclusions about freedom to operate.
Explore filer portfolios in EurekaThis search casts a wide net across hardware, circuit and accelerator terms. A tighter string focused on specific activation-function implementations would narrow the 1,143 families to a more targeted set.
Refine the search in EurekaFilings from 2022 onward are still publishing and will keep adding to their totals for up to 30 months after filing. Re-running this search periodically will fill in those years.
Set up monitoring in EurekaThis search returns 2,369 published records, which collapse to 1,143 simple patent families once the same invention filed in multiple countries is de-duplicated. The search string covers 'activation function' or 'nonlinear function' combined with neural network language and hardware, circuit or accelerator terms, and carries no date limit. Because it is a text-based search, it can also pick up records where those terms appear in an unrelated context, so the totals should be read as what matched the string rather than a curated technology count.
International Business Machine Corporation and Intel Corp are tied at the top of the listed filers, each named on 37 of the 1,143 simple patent families. No single filer holds more than either of them, and the filer list is not rolled up to corporate group, so a parent and its subsidiaries would appear as separate rows rather than combined. A count of families also says nothing about the breadth or enforceability of any filer's claims.
Published records rose from 196 in 2018 to 346 in 2021, a 77% increase over that span, and 2021 is the most recent year that can be treated as complete. Records filed from 2022 onward are still publishing, a filing can take up to 30 months to appear, especially when it enters national phase via a PCT application, so the apparent drop-off in the most recent years is a publication-lag artifact, not a real decline. The chart counts published records, not de-duplicated inventions, so filing the same invention in more countries also raises the count.
G06N, the IPC subclass for computing based on AI models, appears on 900 of the 1,143 families (78.7%), and G06F, electric digital data processing, appears on 396 (34.6%). Because a single family can carry multiple IPC classes, these two shares overlap and add to more than 100% between them. The search returned 100 subclasses in total, but only these two sit at a share worth discussing on this dataset.
1,296 of the 2,369 published records, or 55%, were granted and still in force as of the 2026-10-27 retrieval date. The remainder are pending, withdrawn, refused, lapsed or not determinable, and that breakdown is not separated out in this dataset. In-force status is a snapshot that changes continuously as patents are challenged, lapse or newly grant, so it should not be read as a success rate for the technology.
The United States received the most filings among the offices captured here with 1,161 published records, followed by Europe (EPO) at 274, China at 240, the WIPO PCT route at 213, India at 135 and the United Kingdom at 88. The search returned 24 receiving offices in total, so these six are the largest rather than the complete picture, and the figures are counted in published records rather than de-duplicated families, so one invention filed in several offices adds a count at each one.
Go past this page: query the whole neural network activation function hardware implementation patent landscape corpus yourself, in your own scope.
Every answer comes back with patent numbers you can open.
Search. Every figure on this page comes from one Patsnap search, run once and not edited afterwards. This is the exact search string:
TAC:(("activation function" OR "nonlinear function") AND ("neural network" OR "artificial neuron") AND (hardware OR circuit OR accelerator))
Reading the field codes: TAC searches the title, abstract and claims of each record, not the full description. The search string sets no date limit, so the scope is every matching record in the database regardless of when it was filed. It is not restricted to recent years. How this string was built matters for how far the counts can be pushed. It was generated in batch from a topic list rather than hand-crafted for this page, and the result set is not manually screened. A keyword search over title, abstract and claims both over-captures (unrelated filings that happen to use the same words) and under-captures (the same invention written in different words). This page cannot measure either effect, so read every figure here as what this particular string returned, not as what exists in the field. One subclass (G06N) carries 78.7% of the patent families, so this search returned a classification-tight set.
What is being counted. Two different units appear on this page and they are not interchangeable. The filer list and the technology split are counted in simple (DOCDB) patent families: one invention counts once however many countries it was filed in. The total, the jurisdiction split and the filing-year chart are counted in published records: the same invention counts once per publication. For this search that is 2,369 records and 1,143 families. Record counts reward filing in many jurisdictions; family counts do not. Shares are only ever taken against a denominator in the same unit, and each one names it.
Who the filers are. Names come from Patsnap's standardised assignee field, as recorded on the documents themselves. They are not rolled up to corporate group: a parent and its subsidiaries appear as separate entries, and a portfolio acquired from another company stays under the name on the document unless the assignment is reflected there. This list is meant to carry organisations, not people. Filings held by a named individual are identified by matching the assignee against the inventors on the same search, and, for names written in Chinese, Japanese or Korean characters, by the name carrying no organisational suffix; those are removed from the list (1 of them here). Neither test is exhaustive: the inventor list it matches against is itself capped, and a personal name written in Latin script with no organisational wording in it cannot be told apart from a short company name by rule. So an individually held filing can still appear in the list. Those filings stay inside every total and share on this page either way, so the listed entries can add up to less than the total. A record shown in full further down still names its assignee exactly as the patent office records it, including where that is a person: that name is how the record is identified and checked. What is left out of the ranking is people as ranked entries, not people as cited sources. The table shows the top 30 of the 99 filers returned, ordered by count alone; the data endpoint returns at most 100 entities, so even that 99 is not every filer. One patent family can name several filers and each is counted separately, so these rows are not slices of a divided whole: adding them together over-counts, and this page deliberately publishes no combined "top five" share for that reason. See every filer, individuals included
Dates. The year chart counts by earliest filing year, not publication year. A filing only becomes visible once it publishes: roughly 18 months for most routes, and up to 30 months where a PCT application enters national phase. Measured across these pages, the year four back from today still holds only about 89% of its eventual count. 2022 onwards are therefore shaded on the chart as still filling in, and the growth figure stops at 2021. The data behind this page was retrieved on 2026-10-27.
Legal status. "In force" counts documents whose legal status at retrieval was granted and maintained (1,296 of 2,369). Everything else is pending, withdrawn, refused, lapsed or not determinable, and those states are not broken out here because the underlying codes do not separate them cleanly. Legal status changes continuously and is reported per document, not per family. Get the status breakdown in Eureka
Classification. The technology split uses IPC subclasses (four characters, for example H01M). A single record usually carries several, so the shares add up to more than 100% and a record can appear in more than one bar. Subclass is a coarse grouping; it describes the field a record was classified into, not what the claims cover. Break the classes down to subgroup level
Who wrote the words. The tables, charts, counts, shares and patent numbers on this page are produced directly from the search result by a program; none of them are written or estimated by a language model. The surrounding prose, headings, FAQ answers and the answers in the question panel are drafted by a language model that is given those same figures and is not permitted to introduce others. Prose can still mis-describe a correct figure, so where the two disagree, the tables and charts are the record.
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 limitations that apply to every figure above. Recent filings are under-counted, because an application is not public until roughly 18 months after filing and database loading adds further delay. The search string defines and constrains everything shown: records that use different terminology, or that sit outside the fields searched, are absent and cannot be counted. Applicant names, classification codes, legal-status codes and citation records are taken as recorded and may contain errors, duplicates, omissions or stale entries. Ordering by volume is not a measure of quality, scope, enforceability or commercial value, and a larger portfolio is not a stronger one. 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, validity or infringement search, nor does it constitute legal, financial, investment, or professional advice, and it should not be relied upon as such. No attorney-client relationship is created by it. Any patent, commercial, or strategic decision should be verified independently against the primary records 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.
The question panel. The answers behind the suggested questions on this page were written when the page was generated, not in response to your click. The progress steps that play before an answer appears are an animation of how that analysis was done; nothing is being searched or computed at that moment. Typing your own question instead opens Patsnap Eureka, which does run it live.
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