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 →9 patent families and 12 records cover weight-stationary dataflow in processing element arrays, led by a single filer with 2 families.
12 published records collapse into 9 simple (DOCDB) families; the difference is the same invention filed in more than one country. In force = 5 of 12 records still granted and maintained at the data cut-off; the rest are pending, withdrawn, refused or lapsed.
This page reports a targeted search for weight-stationary dataflow architectures built on arrays of processing elements. The query combines the terms "weight-stationary" (or "weight stationary"), "dataflow", and either "processing element*" or "array" against the title, abstract and claims of each record. No date limit was set, so the result set reflects every matching record the search returns regardless of filing year, not a particular span of years.
The result is a small pool: 9 simple patent families and 12 published records in total. At this size, counts of filers, classifications or filing years describe exactly what the search returned and nothing more. This page is a starting point for a more targeted search, not a map of the field.
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. Filings held by named individuals rather than organisations are removed where they can be identified; 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 8 of the 8 filers returned for this search; the data endpoint returns at most 100 entities, so even that 8 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.
Three views of the same 9-family, 12-record dataset: where applications were filed, how filings are spread across the years on record, and which IPC subclasses recur most.
Records run from a 2018 filing up to 2026, with the 2025 filing year at 5 records, the peak so far. Years from 2022 onward are still filling in as later publications arrive, so the recent slice of this line understates eventual totals; it should not be read as a rise or fall in filing activity.
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.
G06F (electric digital data processing) appears on 7 of the 9 families (77.8%) and G06N (computing based on AI models) on 6 of the 9 (66.7%). A family can carry both classes at once, which is why the two shares add to more than 100%.
Shares are the percentage of the 9 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 processing element array weight-stationary dataflow architecture patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaUS20240013042A1, filed by Taiwan Semiconductor Manufacturing Company, Ltd. and published 2024-01-11, describes a pipelined processing core with two stages: a first processing core with a matrix array of processing elements arranged in columns and rows, each performing a multiply-accumulate operation on an input and a weight, feeding its output into a second processing core built from a column of processing elements performing further multiply-accumulate operations.This is the first record returned by the search under the database's default ordering, not a ranking by citations, grant status or breadth. It illustrates one way the search terms match a filing; it says nothing about the dataset as a whole.


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 →Four figures drawn directly from this 9-family, 12-record search, each with the unit it is counted in and what it does and does not establish.
The search returns 9 simple patent families (de-duplicated inventions) and 12 published records (one per publication). No date limit was applied, so this reflects every matching record the search returns, not a bounded time window.
Northwestern Univ is named on 2 of the 9 families in scope, the highest share held by any single listed filer. The other 7 listed filers each hold 2 or fewer families; filer rows overlap because one family can name more than one filer, so rows are never added together into a combined share.
5 of the 12 published records (42%) were granted and still in force at the 2026-10-21 retrieval date. The remainder cover other statuses that are not broken out here; this figure describes legal status at a moment, not a success rate.
The 12 published records were filed across three receiving offices: the United States (9), India (2) and WIPO under the PCT route (1). This counts where filings were lodged, not where any filer is based.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to processing element array weight-stationary dataflow architecture patent landscape, with the prior art for and against each one.
A 9-family pool is narrow enough that the next useful step is usually a closer read of the documents themselves, not a broader statistical view.
Family and filer counts describe volume, not claim scope. Checking what any one filing actually covers means pulling its granted claims directly.
Open a record in EurekaA 9-family result is sensitive to the exact phrasing used. Testing adjacent terms can reveal whether this pool reflects the full extent of relevant filings or just one slice of the vocabulary used to describe it.
Refine the search in EurekaThis search returns 9 simple patent families and 12 published records with no date limit applied, so it reflects every matching record the search string finds. A simple family de-duplicates the same invention across countries, while published records count each publication separately, so the two numbers are not interchangeable. Because the pool is small, a differently worded search string could return a different count; this figure describes this exact query, not a fixed total for the technology.
Northwestern Univ is named on 2 of the 9 simple patent families in scope (22.2%), the highest share held by any single listed filer. The search returned 8 organisation-level filers in total, each holding 2 or fewer families, so the field shown here is spread across a small set of entities rather than concentrated in one. Filer rows cannot be added together because a single family can name more than one filer, so no combined share for any group of rows can be read from this list.
Its abstract describes a pipelined processing core with two stages: a first stage built from a matrix array of processing elements performing multiply-accumulate operations on an input and a weight, feeding a second stage built from a column of processing elements performing further multiply-accumulate operations. Taiwan Semiconductor Manufacturing Company filed it, and it published on 2024-01-11. An abstract and a family count describe what a filing discusses, not what its granted claims legally require; determining whether any specific design falls inside or outside its protection means reading the granted claim set directly.
Two IPC subclasses recur across the 9 families in scope: G06F (electric digital data processing) on 7 of them, 77.8%, and G06N (computing based on AI models) on 6 of them, 66.7%. These shares add to more than 100% because a single family can carry both classes at once. Other subclasses appear in the underlying data at lower shares and are not detailed here, since this small a pool cannot reliably separate a genuine minor branch from an unrelated classification picked up by broad search terms.
The filing trend recorded by earliest filing year runs from 1 record in 2018 to a peak of 5 records in 2025, with 2 recorded so far in 2026. Years from 2022 onward are still filling in as later publications arrive, recent filings can take 18 to 30 months to publish, so the most recent years understate their eventual totals and should not be read as a slowdown or a rise. With only 12 records spread across these years, year-to-year movement in this chart is too small a sample to support a trend conclusion.
Go past this page: query the whole processing element array weight-stationary dataflow architecture 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:(("weight-stationary" OR "weight stationary") AND dataflow AND ("processing element*" OR array))
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 (G06F) carries 77.8% 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 12 records and 9 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. 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 8 of the 8 filers returned, ordered by count alone; the data endpoint returns at most 100 entities, so even that 8 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. The data behind this page was retrieved on 2026-10-21.
Legal status. "In force" counts documents whose legal status at retrieval was granted and maintained (5 of 12). 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.
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