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/high-performance-storage-nvme-over-fabrics-patent-landscape-patent-landscape/ · Patsnap · data cut-off 2026-08-31 · downloaded from the live page
Patent Landscape · High-Performance StorageNVMe over Fabrics Patents: Who Holds the Claims and Where the Field Is Still Open
A data-led view of the NVMe over Fabrics patent landscape: who holds the concentrated top filings, how filing activity has moved since the 2020 peak, which IPC classes carry the claim density, and where the white space s
808
Published Records
62%
Top-5 Share of All Records
-76%
Filing Growth 2021→2024
US
Leading Jurisdiction
Filing growth = 2021 (131 records) → 2024 (32); 2024 is the last year we treat as complete. Top-5 share = the 5 largest assignees ÷ all 808 records in scope (CR5), not the ranked leaders only.
Published byPatsnap Research··7 min readSourced from Patsnap Eureka
Overview
What the NVMe over Fabrics patent record actually shows
NVMe over Fabrics extends the NVMe command set across a network fabric — RDMA, Fibre Channel, TCP — so a remote drive behaves, from the host's point of view, like a local one. The patent record built around this topic spans 808 published records from 2015 through the current data cut-off, concentrated heavily in storage, networking and server infrastructure filers. Family-level counting is used throughout so that continuation filings and multi-jurisdiction copies of the same invention are not counted twice.
The filing curve rose through the late 2010s, peaked in 2020, and has since declined — a pattern consistent with a core transport-layer technology whose foundational claims were staked early. Publication lags filing by roughly 18 months, so the most recent one or two years in any chart will always look thinner than they will eventually turn out to be; that lag, not a collapse in R&D interest, explains the low counts nearest the present.
Filing activity by year, 2017–2026
Source: Patsnap Eureka. Assignee ranking and totals. Derived from a Patsnap search on High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Let an AI agent run this analysis on your own technology
Pick a task. Every answer cites the patents behind it.
10,000 free credits to start
The data
Filing trends and technology composition
Three views of the same 808-record dataset: when the filings happened, who filed them, and which technical classes they sit in.
A sharp rise, an early peak, a sustained pullback
Filings ran at 75 in 2017 and climbed to a peak of 143 in 2020. From 2021's 131 down to 32 in 2024 — the last year that can be treated as complete — activity fell 76%, a contraction consistent with the core transport-layer claim space having already been staked out by the leaders during the protocol's standardisation years.
Publication lags filing by roughly 18 months, so 2025 onwards are still filling in. Growth rates on this page therefore end at 2024; running them to the last bar would understate the field.
Claim density sits almost entirely in two classes
G06F (electric digital data processing) appears on 85.4% of the 808 records and H04L (digital information transmission) on 35.0% — the expected pairing for a protocol that lives at the intersection of storage command sets and network transport. Every other class, from AI-model computing to printed-circuit assemblies, sits below 2.2%, which is a share worth noting rather than ignoring: it marks where NVMe-oF claims have not yet been pulled into adjacent hardware or AI-adjacent filings in volume.
Shares are the percentage of the 808 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 High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. Counts reflect published records only and shift as new filings publish.
Go deeper on High-Performance Storage: NVMe Over Fabrics Patent Landscape with Eureka
This page is one run against one query. Ask Eureka your own question about high-performance storage: nvme over fabrics patent landscape and every answer comes back with the patent numbers behind it.
Try EurekaKey patents
The most-cited records and a representative filing
Representative filing
US20180067685A1 — Method for Implementing NVMe over Fabrics, Terminal, Server, and System
HUAWEI TECHNOLOGIES CO.,LTD.
A method for implementing NVMe over fabrics includes generating, by a terminal, a NVMe instruction, where the NVMe instruction indicates a data read operation or a data write operation. The method further includes sending, by the terminal by using remote direct memory access (RDMA), the NVMe instruction to a submission queue (SQ) that is stored in a server. When the NVMe instruction indicates the data read operation, the method includes receiving, by the terminal by using the RDMA, to-be-read data sent by the server. Alternatively, when the NVMe instruction indicates the data write operation, the method includes sending, by the terminal, to-be-written data to the server by using the RDMA.Filed by Huawei Technologies, published 2018-03-08 — an early, broad claim to the terminal-to-server RDMA read/write pattern that underlies most subsequent NVMe-oF implementations.
Most-cited records in the dataset
| # | Publication no. | Patent title | Citations |
|---|---|---|---|
| 1 | US20220027051A1 | Data Path Virtualization | 186 |
| 2 | US20150254003A1 | RDMA-SSD dual-port unified memory and network controller | 167 |
| 3 | US20190317901A1 | System and method for optimizing performance of a solid-state drive using a deep neural network | 146 |
| 4 | US20200136996A1 | Offload of storage node scale-out management to a smart network interface controller | 115 |
| 5 | US10162793B1 | Storage adapter device for communicating with network storage | 101 |
| 6 | US20200319812A1 | Intermediary for storage command transfers | 96 |
| 7 | US10310760B1 | Layering communication fabric protocols | 87 |
| 8 | US20210019270A1 | Configuration interface to offload capabilities to a network interface | 74 |
| 9 | US20180032463A1 | System architecture for supporting active pass-through board for multi-mode NMVE over fabrics devices | 74 |
| 10 | US20200293465A1 | Multi-protocol support for transactions | 72 |
Citation counts inside any searched corpus favour older filings simply because they have had longer to accumulate citations — read this table as a map of influence, not of current commercial weight.
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 High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. 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 numbers mean for a filing decision
Four read-outs from the dataset, each pointing at a different practical question: where the field is crowded, where it thinned out, and who to watch.
Concentration
62.4%
of 808 records held by the top 5
The top of the field is tight
The leader alone accounts for 244 records, and the top five combined hold 62.4% of all 808 records in scope. That level of concentration means a new entrant is filing directly into space already claimed by a handful of large storage and networking vendors, not into an open field.
Assignee ranking, 99 companies
Filing trend
-76%
2021 → 2024 (131 → 32 records)
The early land-grab has passed
Filings peaked at 143 in 2020 and dropped 76% from 2021's 131 to 32 in 2024, the last complete year. That trajectory fits a transport protocol whose foundational claims were staked early rather than a technology losing relevance — later years are still filling in due to publication lag.
Filing trend, 2017-2026
Technology mix
85.4%
of records carry a G06F class
Claims sit in two classes, not eight
G06F and H04L between them dominate the classification picture, while classes like G06N (AI models), H04N (pictorial communication) and H01R (connectors) each sit under 2.2% of records. That gap is where cross-domain claims — NVMe-oF tied to AI inference pipelines or physical connector design — remain lightly filed.
IPC composition, 8 subclasses shown
Jurisdiction
505
US-office records, the largest single office
Filing is US-centred with real EPO and China volume
The United States receives the largest share of filings at 505 records, with Europe (73), China (61), Germany (44), WIPO/PCT (34) and Japan (26) forming a meaningful second tier. Any freedom-to-operate check needs to cover at least those first three offices to be credible.
Receiving offices, six shown
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 high-performance storage: nvme over fabrics patent landscape, with the prior art for and against each one.
Co-filing is rare in this dataset
| Assignee | Co-assignee | Shared families |
|---|---|---|
| Intel Corporation | ZHU QIANYING | 1 |
| Intel Corporation | ZHANG ZHIYUAN | 1 |
| Intel Corporation | ZHANG XINXIN | 1 |
| Intel Corporation | WU XIANGBIN | 1 |
| Intel Corporation | SHEN YINGZHE | 1 |
| Intel Corporation | JI HAITAO | 1 |
| Marvell Asia Pte Ltd | Marvell Technology | 1 |
| Marvell Asia Pte Ltd | NOY SHAHAR | 1 |
Only 9 co-assignee pairs appear across the 808 records, and the strongest of them link back to a single major filer working with individual named inventors rather than to joint ventures between companies — collaboration filing is the exception here, not the rule.
Source: Patsnap Eureka. Co-assignee relationships and derived observations. Derived from a Patsnap search on High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. 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 few concrete next steps for a team deciding where to file or where to watch.
Map freedom-to-operate against the top filers
With 62.4% of records held by five companies, a freedom-to-operate review that skips those filers is not credible. Start claim charting against the leader's portfolio before drafting new claims in the transport or command-queue space.
Run a freedom-to-operate check →Target the under-claimed adjacent classes
AI-model computing, pictorial communication and connector hardware each sit under 2.2% of the 808 records. A first claim tying NVMe-oF transport to one of those domains has meaningfully less prior art to design around.
Explore white space in Eureka →Track the next 12–18 months of publications
Because publication lags filing by roughly 18 months, the apparent 2024-2026 decline understates real filing activity. Recheck the trend once the most recent cohort has fully published before drawing conclusions about slowing interest.
Set a filing alert →Source: Patsnap Eureka. Forward-looking reading of the same dataset. Derived from a Patsnap search on High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
FAQ
Common questions about the NVMe over Fabrics patent landscape
The assignee ranking is led by a single company with 244 of the 808 records in scope, and the top five filers together hold 62.4% of all records. That concentration extends to 76.5% once the top ten are counted, so the field is dominated by a small group of storage, networking and server infrastructure vendors rather than spread evenly across many filers. A long tail of smaller filers accounts for the remainder, each holding only a handful of records.
No — filings peaked at 143 records in 2020 and have declined since, falling 76% from 131 records in 2021 to 32 in 2024, the most recent year that can be treated as complete. This does not mean the technology is losing relevance; it is more consistent with the core transport-layer claims having been staked out early in the protocol's standardisation period. Figures for 2025 and 2026 are still understated because publication typically lags filing by around 18 months.
The large majority of records — 85.4% of the 808 in scope — carry a G06F classification for electric digital data processing, and 35.0% also carry H04L for digital information transmission, reflecting the technology's position at the intersection of storage command sets and network transport. Smaller shares touch AI-model computing (G06N), pictorial communication (H04N), printed circuits (H05K) and static memories (G11C), each under 2.2% of records. Because a single record can carry multiple classes, these shares add up to more than 100%.
US20180067685A1, filed by Huawei Technologies and published in 2018, claims a method where a terminal generates an NVMe instruction and sends it via remote direct memory access to a submission queue stored on a server, with the server returning read data or accepting written data over the same RDMA path. This is a foundational terminal-to-server RDMA read/write pattern that sits close to the core mechanics most later NVMe-oF implementations rely on. Anyone building on this pattern should read the granted claim scope closely rather than relying on the abstract alone, since the actual enforceable boundary depends on specific claim language not reproduced here.
The clearest gaps sit in classes that appear on very few of the 808 records: AI-model computing (G06N) at 2.2%, pictorial communication (H04N) at 1.9%, printed circuit assemblies (H05K) at 1.6%, static memories (G11C) at 1.0%, image processing (G06T) at 0.9% and connectors (H01R) at 0.9%. These low shares indicate that NVMe-oF has not yet been heavily claimed in combination with AI inference workloads, video pipelines or physical connector design, which is where a first-mover claim would face the least prior art congestion.
Answers are grounded in the same dataset. Derived from a Patsnap search on High-Performance Storage: NVMe Over Fabrics Patent Landscape covering 2015–2026, data cut-off 2026-08-31. Counts reflect published records only and shift as new filings publish.Run this in Eureka MCP
Research High-Performance Storage: NVMe Over Fabrics Patent Landscape in depth with Eureka
Go past this page: query the whole high-performance storage: nvme over fabrics 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.