Scientific Computing Patents: Top Companies & Filing Trends 2026
- 60.6% concentration. The five leading assignees account for 3,559 of the 5,870 records in scope, a tight cluster above a long tail of single-digit filers.
- Filing cooled but did not collapse. Activity peaked at 628 records in 2019 and eased to 266 by 2024, a -12% move over 2021-2024 — the last span the data can call complete.
- G06F dominates the claim map. 83.6% of records touch electric digital data processing, while AI-based computing (G06N, 8.2%) and transmission (H04L, 7.5%) trail well behind.
Filing growth compares 2021 (302 records) with 2024 (266) — 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 5,870 records in scope (CR5), not by the ranked leaders only.
What this landscape covers
This dataset spans 5,870 published records filed or published between 2015 and the 2026 data cut-off, matched against scientific computing terms combined with implementation signals such as cache coherence, compute kernels, memory access and task parallelism. The search string is deliberately narrow: it isolates records that pair a scientific-computing framing with a concrete parallel-processing or memory mechanism, rather than every record that merely mentions high-performance computing in passing.
Because publication trails filing by roughly 18 months, the most recent one to two years in any trend chart will always look thinner than the underlying filing activity actually was. Treat 2025 and 2026 as still filling in, not as a real decline.
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 5,870 records: how filing volume has moved year over year, and which IPC subclasses carry the claim density.
A 2019 peak followed by a real but partial cooldown
Filings rose from 443 in 2017 to a peak of 628 in 2019, then eased. The only fully comparable recent window, 2021 to 2024, shows a -12% move (302 to 266) — a real softening, but not the freefall a glance at 2025-2026's thin, still-arriving numbers might suggest.
Claim density sits overwhelmingly in G06F
G06F (electric digital data processing) touches 83.6% of the 5,870 records, dwarfing G06N (AI-based computing, 8.2%) and H04L (digital transmission, 7.5%). Smaller classes — G06T, G11C, G06Q, H04N, H03M — each sit under 5%, marking narrower, more specific claim territory rather than core contested ground.
Shares are the percentage of the 5,870 records in scope. A patent can carry several IPC classes, so the shares add up to more than 100%.
Go deeper on Scientific Computing Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about scientific computing patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaRepresentative filing and most-cited prior art
Monitoring Apparatus, Device, Method, and Computer Program and Corresponding System (US20220391248A1)
The filing describes a monitoring apparatus that obtains a first compute kernel to monitor alongside one or more second compute kernels, then instructs control circuitry across a computing device's execution units to run the first kernel on one slice of those units while running the second kernel(s) concurrently on other slices.Filed by Intel Corporation, published 2022-12-08 — a concrete instance of the compute-kernel scheduling and monitoring claims that recur across this dataset.
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US6850252B1 | Intelligent electronic appliance system and method | 4,058 |
| 2 | US6400996B1 | Adaptive pattern recognition based control system and method | 2,342 |
| 3 | US6640145B2 | Media recording device with packet data interface | 1,590 |
| 4 | US5875108A | Ergonomic man-machine interface incorporating adaptive pattern recognition based control system | 1,588 |
| 5 | US20070053513A1 | Intelligent electronic appliance system and method | 1,452 |
| 6 | US7006881B1 | Media recording device with remote graphic user interface | 1,217 |
| 7 | US7813822B1 | Intelligent electronic appliance system and method | 1,198 |
| 8 | US6081750A | Ergonomic man-machine interface incorporating adaptive pattern recognition based control system | 992 |
| 9 | US7982776B2 | SBI motion artifact removal apparatus and method | 883 |
| 10 | US6418424B1 | Ergonomic man-machine interface incorporating adaptive pattern recognition based control system | 776 |
Citation counts reward older, widely-referenced filings inside this corpus — read them as a signal of influence on the field, not as a ranking of current commercial relevance.
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 filing strategy
Three patterns worth acting on before drafting or opposing a claim in this space.
The top of the field is tightly held
Five assignees hold 3,559 of the 5,870 records in scope — 60.6% of the field concentrated at the top, with a long tail of single- and double-digit filers behind them. A new entrant is filing into occupied ground almost everywhere G06F claims sit.
Cooling, not collapsing
The only complete recent comparison — 2021 (302 records) to 2024 (266 records) — shows a -12% move. That is a real pullback from the 2019 peak of 628, but publication lag means 2025-2026 figures are not yet a reliable read on where filing stands today.
One subclass dominates the claim map
G06F (electric digital data processing) appears in 83.6% of the 5,870 records, far ahead of G06N (8.2%) and H04L (7.5%). Since a record can carry several classes, this says claim space in G06F is dense — not that the smaller classes are unimportant.
Co-filing is rare and narrow
Only 10 co-assignee pairs surface in the data, and the strongest link — Ould Ahmed Vall Elmoustapha with Valentine Robert at 101 shared records — is an individual-inventor pairing rather than a cross-company alliance.
Eureka can read the same corpus for gaps instead of for coverage: under-claimed branches adjacent to scientific computing patent landscape, with the prior art for and against each one.
Who is filing, and where momentum has gone quiet
The ranked leaders account for the bulk of activity, but recent-year momentum has softened even for the largest filers — worth checking before assuming any single company still leads on current output.
A single leader well ahead of the field
The top-ranked assignee holds 3,070 records, more than five times the fifth-place holder's 65 — an unusually steep drop-off that marks this as a leader-and-long-tail field rather than a multi-way contest.
Even top filers have gone quiet lately
Every named assignee tracked for recent-year momentum shows a steep year-on-year drop, several down to zero filings in the latest year. Given publication lag, this likely overstates the slowdown, but it is consistent across the leaders rather than isolated to one firm.
Filing is concentrated in the US, with real EPO and PCT volume
The United States receives 3,411 records, well ahead of the EPO (933) and WIPO/PCT route (565). Germany, Austria and India each show meaningful but smaller receiving-office volumes, marking secondary but non-trivial filing corridors.
| Assignee | Recent year | YoY |
|---|---|---|
| SambaNova Systems | 6 | -80% |
| Oracle International Corporation | 3 | -83% |
| Intel Corporation | 0 | -100% |
| OULD AHMED VALL ELMOUSTAPHA | 0 | — |
| VALENTINE ROBERT | 0 | — |
| International Business Machines Corporation (IBM) | 0 | -100% |
| CORBAL JESUS | 0 | — |
| REALTIME DATA LLC | 0 | — |
Where to take this analysis
The dataset points to a field with a dominant leader, a cooling but not collapsing filing rate, and dense claim territory in one IPC subclass. The next steps depend on whether you are filing, opposing, or scouting for licensing.
Check freedom-to-operate against the leader's portfolio
With one assignee holding over half the top-5 concentration, any new compute-kernel or memory-access filing should be checked against that portfolio specifically, not just the field in general.
Explore assignee portfolios in EurekaMap the under-claimed sub-areas before drafting
Cache-coherence and cross-kernel scheduling mechanisms show lighter claim density than core G06F territory — a narrower, well-drafted claim there faces less prior art.
Run a white-space search in EurekaCommon questions about this landscape
One assignee leads the ranked field with 3,070 of the 5,870 records in scope, well ahead of the fifth-place holder at 65 and tenth place at 45. This is a leader-and-long-tail structure rather than a close multi-way race, so freedom-to-operate checks should weight that single portfolio heavily. The ranking covers 100 companies total, not a top-50 or top-100 cut, so smaller filers further down the list are still meaningfully represented.
Filing peaked at 628 records in 2019 and the only fully comparable recent window, 2021 to 2024, shows a -12% move from 302 to 266 records. That is a genuine cooldown from the peak, not a sharp collapse. Figures for 2025 and 2026 look thin, but that is expected: publication typically lags actual filing by around 18 months, so those years are still filling in rather than reflecting a real drop-off.
The overwhelming majority — 83.6% of the 5,870 records — touch G06F, electric digital data processing, covering the core compute-kernel, memory-access and parallel-processing mechanisms the search targets. AI-based computing (G06N) and digital transmission (H04L) follow at 8.2% and 7.5% respectively, with image processing, static memory, business-process computing, video communication and coding/conversion classes each below 5%. Because a single record can carry several IPC classes, these shares add up to more than 100% and should not be read as mutually exclusive categories.
The smaller IPC subclasses — G11C (static memory), H03M (coding and conversion), and H04N (pictorial communication) — each cover under 5% of the 5,870 records, suggesting lighter claim density than the crowded G06F territory. Specific mechanisms like cache-coherence protocols for heterogeneous kernels and cross-kernel memory-access scheduling show up as under-claimed relative to the volume of general compute-kernel filings. A narrower claim targeting one of these specific mechanisms is likely to face less prior art than a broad compute-kernel or parallel-processing claim filed against the G06F core.
The United States receiving office accounts for 3,411 records, by far the largest single corridor, followed by the European Patent Office at 933 and the WIPO/PCT route at 565. Germany (214), Austria (168) and India (151) each carry smaller but non-trivial volumes, marking secondary filing corridors worth watching for regional strategy. Applicants weighing where to file first should note the US concentration reflects both market size and where the leading assignees are headquartered.
Research Scientific Computing Patent Landscape in depth with Eureka
Go past this page: query the whole scientific 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.