# CEVA INC (CEVA)

Informational only - not investment advice.

CIK: 0001173489
SIC: 7370 Services-Computer Programming, Data Processing, Etc.
SIC breadcrumb: [Services](/division/I/) > [Business Services](/major-group/73/) > [SIC 7370 Services-Computer Programming, Data Processing, Etc.](/industry/7370/)
Latest 10-K filed: 2026-02-27
SEC page: https://www.sec.gov/edgar/browse/?CIK=1173489
Filing source: https://www.sec.gov/Archives/edgar/data/1173489/000143774926006091/ceva20251231_10k.htm

## At a glance

FY2025 · period end 2025-12-31 · filed 2026-02-27 · accession 0001437749-26-006091 · source: https://data.sec.gov/api/xbrl/companyfacts/CIK0001173489.json

| Metric | Value | FY | Provenance |
| --- | ---: | ---: | --- |
| Revenue | 109,598,000 USD | 2025 | verified |
| Net income | -10,638,000 USD | 2025 | verified |
| Assets | 388,256,000 USD | 2025 | verified |
| Free cash flow | -6,277,000 USD | 2025 | computed |
| Net margin | -9.71% | 2025 | computed |
| Operating margin | -10.35% | 2025 | computed |
| Revenue YoY | +2.49% | 2025 | computed |
| ROE | -3.16% | 2025 | computed |

Computed values are grepcent-computed from the verified facts above and may differ from ratios the company itself reports. Free cash flow = operating cash flow − capital expenditures. Net margin = net income ÷ revenue. Operating margin = operating income ÷ revenue. Revenue YoY = FY2025 revenue ÷ FY2024 revenue − 1 (consecutive fiscal years only). ROE = net income ÷ period-end stockholders' equity.

No market price, no rating, no forecast on this site. Not investment advice.

### Peer percentile fingerprint

| Ratio | CEVA | Peer median | Percentile | N |
| --- | ---: | ---: | ---: | ---: |
| Net margin | -9.7% | 2.4% | 20 | 31 |
| Operating margin | -10.4% | 4.2% | 23 | 31 |
| Revenue growth | 2.5% | 5.5% | 29 | 32 |
| FCF margin | -5.7% | 15.0% | 10 | 31 |
| ROE | -3.2% | 4.5% | 38 | 27 |
| ROA | -2.7% | 1.6% | 35 | 32 |
| Liabilities / equity | 0.15 | 1.05 | 4 | 27 |
| Current ratio | 9.93 | 1.98 | 97 | 32 |

Percentile = share of the N covered peers reporting that ratio whose value is lower (ties counted half); computed among grepcent-covered companies in SIC industry 7370 Services-Computer Programming, Data Processing, Etc., not the whole market. A higher percentile means a higher value of the ratio, not a better company. Ratios with fewer than 8 reporting peers are omitted. Latest reported values per company; fiscal periods may differ. Descriptive arithmetic - not a score, rating, or ranking.

## Selected Fundamentals
| Metric | Value | Unit | FY | Filed |
| --- | ---: | --- | ---: | --- |
| Revenue | 109598000 | USD | 2025 | 2026-02-27 |
| Net income | -10638000 | USD | 2025 | 2026-02-27 |
| Assets | 388256000 | USD | 2025 | 2026-02-27 |

## Financials

Annual standardized facts from SEC companyfacts as of latest extracted filing date 2026-02-27. Source: https://data.sec.gov/api/xbrl/companyfacts/CIK0001173489.json. Derived margins, ratios, and free cash flow are computed from the extracted annual SEC facts.

| Metric | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Revenue | 72,653,000 | 87,507,000 | 77,877,000 | 87,152,000 | 100,326,000 | 113,832,000 | 120,583,000 | 97,419,000 | 106,939,000 | 109,598,000 |
| Net income | 13,100,000 | 17,028,000 | 574,000 | 28,000 | -2,379,000 | 396,000 | -23,183,000 | -11,878,000 | -8,786,000 | -10,638,000 |
| Operating income | 14,386,000 | 15,873,000 | -1,245,000 | -1,924,000 | -763,000 | 7,040,000 | 3,896,000 | -13,467,000 | -7,545,000 | -11,346,000 |
| Gross profit | 66,567,000 | 80,554,000 | 69,926,000 | 77,046,000 | 89,577,000 | 103,454,000 | 105,452,000 | 85,771,000 | 94,171,000 | 95,440,000 |
| Diluted EPS | 0.61 | 0.75 | 0.03 | 0.00 | -0.11 | 0.02 | -1.00 | -0.51 | -0.37 | -0.44 |
| Operating cash flow | 14,459,000 | 24,469,000 | 8,612,000 | 9,674,000 | 15,163,000 | 25,804,000 | 6,924,000 | -6,331,000 | 3,471,000 | -3,356,000 |
| Capital expenditures | 2,387,000 | 4,135,000 | 3,319,000 | 3,461,000 | 2,935,000 | 2,193,000 | 3,499,000 | 2,884,000 | 2,955,000 | 2,921,000 |
| Share buybacks | 3,417,000 | 0.00 | 20,008,000 | 9,113,000 | 4,780,000 | 0.00 | 6,785,000 | 6,163,000 | 8,456,000 | 7,150,000 |
| Assets | 242,495,000 | 276,812,000 | 277,263,000 | 297,021,000 | 306,952,000 | 328,659,000 | 308,442,000 | 304,085,000 | 308,948,000 | 388,256,000 |
| Stockholders' equity | 211,551,000 | 244,670,000 | 245,879,000 | 251,157,000 | 260,889,000 | 276,732,000 | 258,871,000 | 264,341,000 | 266,556,000 | 336,458,000 |
| Cash and cash equivalents | 18,401,000 | 21,739,000 | 22,260,000 | 22,803,000 | 21,143,000 | 32,642,000 | 20,116,000 | 23,287,000 | 18,498,000 | 40,586,000 |
| Free cash flow | 12,072,000 | 20,334,000 | 5,293,000 | 6,213,000 | 12,228,000 | 23,611,000 | 3,425,000 | -9,215,000 | 516,000 | -6,277,000 |

### Ratios

ROE and ROA use period-end equity/assets. Liabilities / equity uses total liabilities divided by stockholders' equity. Current ratio uses current assets divided by current liabilities when both are reported.

| Metric | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 |
| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
| Net margin | 18.03% | 19.46% | 0.74% | 0.03% | -2.37% | 0.35% | -19.23% | -12.19% | -8.22% | -9.71% |
| Operating margin | 19.80% | 18.14% | -1.60% | -2.21% | -0.76% | 6.18% | 3.23% | -13.82% | -7.06% | -10.35% |
| Return on equity | 6.19% | 6.96% | 0.23% | 0.01% | -0.91% | 0.14% | -8.96% | -4.49% | -3.30% | -3.16% |
| Return on assets | 5.40% | 6.15% | 0.21% | 0.01% | -0.78% | 0.12% | -7.52% | -3.91% | -2.84% | -2.74% |
| Liabilities / equity | 0.15 | 0.13 | 0.13 | 0.18 | 0.18 | 0.19 | 0.19 | 0.15 | 0.16 | 0.15 |
| Current ratio | 6.40 | 6.98 | 8.15 | 6.77 | 5.95 | 5.33 | 5.34 | 7.79 | 7.09 | 9.93 |

## As-reported value updates

7 tracked differences above grepcent's stated thresholds were found between the earliest XBRL-filed value and the value currently on file for the same fiscal period.

Ledger: /company/CEVA/revisions/


## Quarterly

Quarterly standardized facts from SEC companyfacts as of latest extracted filing date 2026-08-10. Source: https://data.sec.gov/api/xbrl/companyfacts/CIK0001173489.json.

Flow metrics use discrete quarter-length periods from 10-Q/10-Q/A filings. Q4 revenue and net income are derived only when annual FY and nine-month YTD facts exist for the same fiscal year; derived Q4 values are labeled. EPS Q4 is not derived.

| Quarter | End date | Revenue | Net income | Diluted EPS | Method |
| --- | --- | ---: | ---: | ---: | --- |
| 2018-Q2 | 2018-06-30 | 17,494,000 |  |  | reported discrete quarter |
| 2018-Q3 | 2018-09-30 | 21,413,000 |  |  | reported discrete quarter |
| 2019-Q2 | 2019-06-30 |  |  | -0.07 | reported discrete quarter |
| 2019-Q3 | 2019-09-30 |  |  | 0.03 | reported discrete quarter |
| 2020-Q1 | 2020-03-31 |  |  | -0.05 | reported discrete quarter |
| 2020-Q2 | 2020-06-30 |  |  | -0.05 | reported discrete quarter |
| 2020-Q3 | 2020-09-30 |  |  | -0.03 | reported discrete quarter |
| 2021-Q1 | 2021-03-31 |  |  | -0.16 | reported discrete quarter |
| 2021-Q2 | 2021-06-30 |  |  | 0.01 | reported discrete quarter |
| 2021-Q3 | 2021-09-30 |  |  | -0.01 | reported discrete quarter |
| 2022-Q1 | 2022-03-31 |  |  | -0.07 | reported discrete quarter |
| 2022-Q2 | 2022-06-30 |  |  | -0.05 | reported discrete quarter |
| 2022-Q3 | 2022-09-30 |  |  | -0.96 | reported discrete quarter |
| 2023-Q1 | 2023-03-31 |  |  | -0.21 | reported discrete quarter |
| 2023-Q3 | 2023-09-30 |  | -4,957,000 |  | reported discrete quarter |
| 2023-Q4 | 2023-12-31 |  | 3,769,000 |  | derived Q4 = FY annual - nine-month YTD |
| 2024-Q1 | 2024-03-31 |  | -5,448,000 |  | reported discrete quarter |
| 2024-Q2 | 2024-06-30 |  | -291,000 |  | reported discrete quarter |
| 2024-Q3 | 2024-09-30 |  | -1,311,000 |  | reported discrete quarter |
| 2024-Q4 | 2024-12-31 |  | -1,736,000 |  | derived Q4 = FY annual - nine-month YTD |
| 2025-Q1 | 2025-03-31 |  | -3,327,000 |  | reported discrete quarter |
| 2025-Q2 | 2025-06-30 |  | -3,704,000 |  | reported discrete quarter |
| 2025-Q3 | 2025-09-30 |  | -2,509,000 |  | reported discrete quarter |
| 2025-Q4 | 2025-12-31 |  | -1,098,000 |  | derived Q4 = FY annual - nine-month YTD |
| 2026-Q1 | 2026-03-31 | 27,024,000 | -4,459,000 |  | reported discrete quarter |
| 2026-Q2 | 2026-06-30 | 29,033,000 | -2,910,000 |  | reported discrete quarter |

## Filed narrative (10-K & 10-Q)

## Business

Verbatim Item 1 Business section from CEVA's latest 10-K: [/company/CEVA/business/](/company/CEVA/business/).

## Risk Factors

Verbatim Item 1A Risk Factors from CEVA's latest 10-K: [/company/CEVA/risk-factors/](/company/CEVA/risk-factors/).

## Latest quarter (10-Q)

Latest 10-Q source: https://www.sec.gov/Archives/edgar/data/1173489/000143774926026774/ceva20260630_10q.htm

Extracted structurally from real Item 2 body heading to real Item 3/4 boundary. Published MD&A gate trimmed front/tail over-capture.
Confidence: high
Filing date: 2026-08-10
Report date: 2026-06-30

Item 2. MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS

You should read the following discussion together with the unaudited financial statements and related notes appearing elsewhere in this Quarterly Report. This discussion contains forward-looking statements that involve risks and uncertainties. Any or all of our forward-looking statements in this Quarterly Report may turn out to be wrong. These forward-looking statements can be affected by inaccurate assumptions we might make or by known or unknown risks and uncertainties. We undertake no obligation to publicly update any forward-looking statements, whether as a result of new information, future events or otherwise. Factors which could cause actual results to differ materially include those set forth in Part II—Item 1A—“Risk Factors” in this Quarterly Report and Part I—Item 1A—“Risk Factors” in our Annual Report on Form 10-K for the year ended December 31, 2025 as well as those discussed elsewhere in this Quarterly Report. See “Forward-Looking Statements.”

The financial information presented in this Quarterly Report includes the results of Ceva, Inc. and its subsidiaries.

BUSINESS OVERVIEW

We enable Physical AI, the artificial intelligence embedded in billions of devices that connect, sense and infer data in the real world. We view Physical AI as the natural evolution of Edge AI. While Edge AI refers to running AI workloads locally on devices rather than in the cloud, Physical AI extends this concept further: it unifies connectivity, sensing and inference layers into a single fabric that allows devices not only to process data at the edge, but also to interact intelligently with their physical environment and the cloud. We believe Ceva is uniquely positioned as the only company with leadership in innovative silicon and software IP solutions across all three layers – connect, sense and infer.

In the second quarter of 2026, we continued to execute on our strategy, generating revenues of $29.0 million, up 13% year-over-year. Licensing and related revenues were $18.2 million, up 21% year-over-year and representing our strongest quarterly licensing performance in more than three years. Royalty revenues were $10.8 million, up 17% sequentially, reflecting improving smartphone royalties, continued strength in wireless connectivity and the ongoing ramp of automotive AI programs.

According to IPnest, we commanded 68% of the wireless connectivity IP market in 2024. Since 2003, more than 20 billion devices have shipped with Ceva IP, including approximately 2.1 billion in 2025. Our technologies power the connectivity, perception and intelligence in today’s most advanced smart edge products across consumer IoT, automotive, industrial and infrastructure, and mobile and PC markets. Based on market research, we believe these sectors will represent a $170 billion total addressable market for Physical AI and Edge AI by 2030.

Our portfolio spans:

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer (wireless transport): Bluetooth, Wi\u2011Fi, Ultra\u2011Wideband (UWB), cellular internet-of-things (IoT), and 5G\u2011Advanced IP platforms that form the backbone of ubiquitous, secure, and high\u2011performance communication."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer (software and DSPs): Sensor fusion processors, RealSpace spatial audio, MotionEngine software, and general\u2011purpose digital signal processors (DSPs) that transform raw sensor data into actionable intelligence."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer (NPUs and AI DSPs): The NeuPro family of neural processing units (NPUs), from NeuPro\u2011Nano for embedded AI to NeuPro\u2011M for generative AI, supported by a unified toolchain and software stack for simple model deployment, and the SensPro family of AI DSPs for high\u2011performance signal and AI workloads."]]
[[/GREPCENT_TABLE]]

Together, these layers make Ceva, with our unified AI fabric, an essential enabler of Physical AI that breaks down barriers to entry and accelerates time‑to‑market for our customers. We are increasingly delivering more integrated, system-level solutions, rather than individual IP blocks. This approach enables customers to accelerate development, reduce engineering risk and focus internal resources on system-level differentiation, while increasing Ceva’s content per platform, deepening customer relationships and strengthening long-term royalty potential.

For more than three decades, we have been a trusted partner to hundreds of leading semiconductor and original equipment manufacturer (OEM) companies, serving not only our largest target growth markets but also a wide variety of other applications, including smart home, surveillance, robotics and medical. Our transformative semiconductor IP and embedded software offerings are incorporated by customers into application-specific integrated circuits (ASICs) and application-specific standard products (ASSPs) to enable power‑efficient, intelligent, secure and connected devices that connect, sense and infer – the three critical pillars of the rapidly evolving era of AI‑enabled smart edge technology.

We license our portfolio of wireless communications, sensing and scalable Edge AI IP to our customers, breaking down barriers to entry and enabling them to bring new cutting-edge products to market faster, more reliably, efficiently and economically.

28

Table of Contents

We believe our portfolio of technologies comprised of connectivity, sensing and inference – the three foundational layers of Physical AI – positions Ceva at the center of the most important megatrends shaping the semiconductor industry, including 5G expansion, generative and embedded AI, industrial automation and vehicle electrification. Demand across these areas continues to drive strong interest in our IP portfolio, both in established markets and in new, emerging use cases.

In the second quarter of 2026, we signed 10 IP licensing agreements, including two with first-time customers and two directly with OEMs. The quarter included one of the most strategically significant AI licensing agreements in our history, with a leading global AI and computing platform company selecting our NeuPro-M NPU IP for its next-generation custom AI silicon. This engagement represents a new category of customer for Ceva and enables close collaboration across both the hardware accelerator and AI software stack, allowing the complete AI pipeline to be optimized for the customer’s specific models, applications and use cases.

The quarter also reflected increased adoption of our diverse portfolio of broader connectivity solutions. A high-volume U.S. semiconductor company adopted, as part of its own product portfolio, a chip based on our Wi-Fi 6 and Bluetooth Low Energy IP that had been developed with another Ceva customer. Separately, an existing customer expanded from licensing a single baseband component to adopting our complete baseband processing subsystem. These engagements illustrate how customers are increasingly leveraging Ceva for integrated solutions that accelerate time-to-market and reduce development risk.

The remaining licensing agreements were primarily across our connectivity portfolio, with customer wins in Europe, China and across Asia-Pacific, demonstrating the broad-based and global nature of demand for our wireless technologies.

AI-related licensing represented more than 20% of total licensing and related revenues in the quarter, highlighting the accelerating adoption of edge AI across multiple end markets.

We believe the following key elements represent significant growth drivers for Ceva as the leader in silicon and software IP enabling Physical AI, spanning the three foundational layers of connectivity, sensing and inference:

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 Foundation for billions of devices: Our broad Bluetooth, Wi\u2011Fi, UWB, and cellular IoT IP platforms address the high\u2011volume IoT, industrial, consumer, and smart home markets. ABI Research projects more than 16.5 billion devices annually by 2029. With leadership in Wi\u2011Fi 6 and Wi\u2011Fi 7 IP, and record Wi\u2011Fi 6 shipments in 2025, we believe we are positioned to capture higher royalty revenues as customers transition to newer standards."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 5G everywhere: Our PentaG2 platform and DSPs for 5G mobile broadband, 5G RedCap and 5G Advanced, including our PentaG-NTN solution for satellite communications, provide one of the industry\u2019s most comprehensive baseband IP solutions, enabling fixed wireless access, satellite communications, robotics, automotive and industrial applications."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 Infrastructure intelligence: Our PentaG RAN platform, including the Ceva\u2011XC22 multi\u2011thread DSP, extends our leadership into 5G RAN and 5G Advanced for data centers and infrastructure, enabling scalable, customizable solutions for next\u2011generation networks."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer \u2013 Consumer audio and spatial intelligence: High\u2011volume consumer audio markets such as TWS earbuds, AR/VR headsets, and wearables represent incremental growth opportunities for our Bluetooth, Audio AI DSPs, NPUs and RealSpace Spatial Audio & Head Tracking software. Recent design wins with Nothing and other consumer brands highlight growing adoption."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer \u2013 Software intelligence at scale: Our MotionEngine software, already shipped in more than 500 million devices, enhances MEMS\u2011based inertial and environmental sensors across robotics, smartphones, laptops, TWS earbuds, and more. Combined with our SensPro DSPs and NeuPro NPUs, we believe this positions Ceva as a one\u2011stop shop for sensor processing and AI\u2011driven user experiences."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 Generative and classic AI at the edge: Our NeuPro\u2011M AI NPU family delivers efficient, high\u2011performance architectures for generative and classic AI across devices from gateways and notebooks to AR/VR and smartphones. Recent agreements include Microchip\u2019s portfolio license win for our NeuPro-M and NeuPro-Nano family of NPUs, together with a licensing agreement with a leading global AI and computing platform company developing next-generation custom AI silicon. These engagements demonstrate the growing adoption of our AI technologies across semiconductor, OEM and platform customers."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 Embedded AI and TinyML: Our NeuPro\u2011Nano NPUs bring cost\u2011 and power\u2011efficient AI to microcontrollers and systems-on-chips (SoCs), enabling artificial IoT (AIoT) devices for sound, vision, vibration, and health monitoring. ABI Research projects that by 2030, over 50% of TinyML shipments will be powered by dedicated embedded AI hardware such as NeuPro\u2011Nano."]]
[[/GREPCENT_TABLE]]

29

Table of Contents

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 AI DSPs for perception and sensor intelligence: Our SensPro2 AI DSP family addresses demand for efficient, high\u2011performance AI signal processing across sensor\u2011rich applications, from smartphones and drones to automotive ADAS and industrial IoT."]]
[[/GREPCENT_TABLE]]

As a result of our focus on silicon and software IP solutions spanning the connectivity, sensing and inference layers of Physical AI, we believe Ceva is well positioned for sustained, long‑term growth in both shipments and royalty revenues. Our diversified royalty streams reflect a broad range of advanced semiconductor packages (ASPs) – from high‑volume Bluetooth and Wi‑Fi connectivity platform

[Excerpt truncated for page length; source filing is linked above.]

## Latest 10-K MD&A (excerpt)

Latest 10-K Item 7 source: https://www.sec.gov/Archives/edgar/data/1173489/000143774926006091/ceva20251231_10k.htm
Complete FY 2025 MD&A: /company/CEVA/mda/fy2025/

Extracted structurally from real Item 7 body heading to real Item 7A/8 boundary.
Confidence: high
Filing date: 2026-02-27
Report date: 2025-12-31

ITEM 7. MANAGEMENT’S DISCUSSION AND ANALYSIS OF FINANCIAL CONDITION AND RESULTS OF OPERATIONS

You should read the following discussion together with the consolidated financial statements and related notes appearing elsewhere in this annual report. This discussion contains forward-looking statements that involve risks and uncertainties. Actual results may differ materially from those included in such forward-looking statements. Factors that could cause actual results to differ materially include those set forth under “Risk Factors,” as well as those otherwise discussed in this section and elsewhere in this annual report. See “Forward-Looking Statements and Industry Data.”

BUSINESS OVERVIEW

We enable Physical AI, the artificial intelligence embedded in billions of devices that connect, sense and infer data in the real world. We view Physical AI as the natural evolution of Edge AI. While Edge AI refers to running AI workloads locally on devices rather than in the cloud, Physical AI extends this concept further: it unifies connectivity, sensing and inference layers into a single fabric that allows devices not only to process data at the edge, but also to interact intelligently with their physical environment and the cloud. We believe Ceva is uniquely positioned as the only company with leadership in innovative silicon and software IP solutions across all three layers – connect, sense and infer.

According to IPnest, we commanded 68% of the wireless connectivity IP market in 2024. Since 2003, more than 20 billion devices have shipped with Ceva IP, including approximately 2.1 billion in 2025. Our technologies power the connectivity, perception and intelligence in today’s most advanced smart edge products across consumer IoT, automotive, industrial and infrastructure, and mobile and PC markets. Based on market research, we believe these sectors will represent a $170 billion total addressable market for Physical AI and Edge AI by 2030.

Our portfolio spans:

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer (wireless transport): Bluetooth, Wi\u2011Fi, Ultra\u2011Wideband (UWB), cellular internet-of-things (IoT), and 5G\u2011Advanced IP platforms that form the backbone of ubiquitous, secure, and high\u2011performance communication."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer (software and DSPs): Sensor fusion processors, RealSpace spatial audio, MotionEngine software, and general\u2011purpose digital signal processors (DSPs) that transform raw sensor data into actionable intelligence."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer (NPUs and AI DSPs): The NeuPro family of neural processing units (NPUs), from NeuPro\u2011Nano for embedded AI to NeuPro\u2011M for generative AI, supported by a unified toolchain and software stack for simple model deployment, and the SensPro family of AI DSPs for high\u2011performance signal and AI workloads."]]
[[/GREPCENT_TABLE]]

Together, these layers make Ceva, with our unified AI fabric, an essential enabler of Physical AI that breaks down barriers to entry and accelerates time‑to‑market for our customers.

For more than three decades, we have been a trusted partner to hundreds of leading semiconductor and original equipment manufacturer (OEM) companies, serving not only our largest target growth markets but also a wide variety of other applications, including smart home, surveillance, robotics and medical. Our transformative semiconductor IP and embedded software offerings are incorporated by customers into application-specific integrated circuits (ASICs) and application-specific standard products (ASSPs) to enable power‑efficient, intelligent, secure and connected devices that connect, sense and infer – the three critical pillars of the rapidly evolving era of AI‑enabled smart edge technology.

We license our portfolio of wireless communications and scalable Edge AI IP to our customers, breaking down barriers to entry and enabling them to bring new cutting-edge products to market faster, more reliably, efficiently and economically.

We believe our portfolio of technologies comprised of connectivity, sensing and inference – the three foundational layers of Physical AI – positions Ceva at the center of the most important megatrends shaping the semiconductor industry, including 5G expansion, generative and embedded AI, industrial automation and vehicle electrification. Demand across these areas continues to drive strong interest in our IP portfolio, both in established markets and in new, emerging use cases. In the fourth quarter of 2025, we signed 18 IP licensing agreements that underscore this momentum across connect, sense and infer: three NPU licensing agreements, highlighted by an engagement with a global PC and smart-device leader developing its next-generation AI personal compute architecture and selecting our NeuPro NPU IP; multiple connectivity wins reflecting customer upgrade cycles, including Wi-Fi 7 and combo connectivity agreements; and a meaningful software engagement with a leading TV platform integrating our MotionEngine technology into its smart TV operating system.

36

Table of Contents

Overall in 2025, we signed 54 IP licensing agreements that underscore our momentum across connect, sense and infer: ten NPU licensing agreements, including a comprehensive portfolio license with Microchip to adopt our full NeuPro NPU family across its roadmap and embed AI capabilities across product lines; multiple agreements for our AI DSPs and accelerators across consumer and automotive applications; and close to thirty connectivity agreements for our Bluetooth and Wi-Fi IPs, including Wi-Fi 7 and Bluetooth High Data Throughput wins with major customers adopting next-generation connectivity standards for upcoming roadmaps. These agreements not only validate the breadth of our portfolio and the strength of our multi-IP strategy, but also support multi-year customer roadmaps and volume ramps, reinforcing Ceva’s role as the enabler of Physical AI across consumer, automotive, industrial and compute markets. Strategically, the licensing agreements we signed during 2025 are building long-term royalty trajectory and visibility. Based on these signed agreements and our insight into customer roadmaps, we estimate that they represent an aggregate lifetime royalty potential of $125 million over their expected product lives. While this value will be realized over multiple years and is dependent on customer deployment and market adoption, the magnitude of this opportunity relative to our current royalty base underscores the strength, durability and accelerating momentum of the licensing and royalty flywheel we are building.

We believe the following key elements represent significant growth drivers for Ceva as the leader in silicon and software IP enabling Physical AI, spanning the three foundational layers of connectivity, sensing and inference:

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 Foundation for billions of devices: Our broad Bluetooth, Wi\u2011Fi, UWB, and cellular IoT IP platforms address the high\u2011volume IoT, industrial, consumer, and smart home markets. ABI Research projects more than 16.5 billion devices annually by 2029. With leadership in Wi\u2011Fi 6 and Wi\u2011Fi 7 IP, and record Wi\u2011Fi 6 shipments in 2025, we believe we are positioned to capture higher royalty revenues as customers transition to newer standards."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 5G everywhere: Our PentaG2 platform and DSPs for 5G mobile broadband and 5G RedCap provide one of the industry\u2019s most comprehensive baseband IP solutions, enabling fixed wireless access, satellite communications, robotics, automotive and industrial applications."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Connectivity layer \u2013 Infrastructure intelligence: Our PentaG RAN platform, including the Ceva\u2011XC22 multi\u2011thread DSP, extends our leadership into 5G RAN and 5G Advanced for data centers and infrastructure, enabling scalable, customizable solutions for next\u2011generation networks."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer \u2013 Consumer audio and spatial intelligence: High\u2011volume consumer audio markets such as TWS earbuds, AR/VR headsets, and wearables represent incremental growth opportunities for our Bluetooth, Audio AI DSPs, NPUs and RealSpace Spatial Audio & Head Tracking software. Recent design wins with Nothing and other consumer brands highlight growing adoption."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Sensing layer \u2013 Software intelligence at scale: Our MotionEngine software, already shipped in more than 500 million devices, enhances MEMS\u2011based inertial and environmental sensors across robotics, smartphones, laptops, TWS earbuds, and more. Combined with our SensPro DSPs and NeuPro NPUs, we believe this positions Ceva as a one\u2011stop shop for sensor processing and AI\u2011driven user experiences."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 Generative and classic AI at the edge: Our NeuPro\u2011M AI NPU family delivers efficient, high\u2011performance architectures for generative and classic AI across devices from gateways and notebooks to AR/VR and smartphones. Recent agreements include Microchip\u2019s portfolio license win for our NeuPro-M and NeuPro-Nano family of NPUs."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 Embedded AI and TinyML: Our NeuPro\u2011Nano NPUs bring cost\u2011 and power\u2011efficient AI to microcontrollers and systems-on-chips (SoCs), enabling artificial IoT (AIoT) devices for sound, vision, vibration, and health monitoring. ABI Research projects that by 2030, over 50% of TinyML shipments will be powered by dedicated embedded AI hardware such as NeuPro\u2011Nano."]]
[[/GREPCENT_TABLE]]

[[GREPCENT_TABLE]]
[["","\u25cf","Inference layer \u2013 AI DSPs for perception and sensor intelligence: Our SensPro2 AI DSP family addresses demand for efficient, high\u2011performance AI signal processing across sensor\u2011rich applications, from smartphones and drones to automotive ADAS and industrial IoT. Bloomberg Intelligence forecasts $58 billion in computer vision AI hardware and $110 billion in conversational AI hardware revenues by 2032, underscoring the scale of this opportunity. SensPro, as an inference\u2011class AI DSP, extends Ceva\u2019s leadership beyond NPUs into versatile, programmable AI compute."]]
[[/GREPCENT_TABLE]]

As a result of our focus on silicon and software IP solutions spanning the connectivity, sensing and inference layers of Physical AI, we believe Ceva is well positioned for sustained, long‑term growth in both shipments and royalty revenues. Our diversified royalty streams reflect a broad range of advanced semiconductor packages (ASPs) – from high‑volume Bluetooth and Wi‑Fi connectivity platforms that power billions of consumer devices to higher‑value inference engines and AI DSPs such as NeuPro and SensPro, as well as infrastructure‑class platforms like PentaG-RAN. We believe this mix provides both scale and resilience, enabling us to capture growth across consumer, automotive, industrial and infrastructure markets while reinforcing our role as the enabler of Physical AI.

37

Table of Contents

CURRENT TRENDS

We believe the long-term trend of digital transformation is evolving into a new era defined by Physical AI – the next phase of Edge AI – where intelligence is embedded directly into the devices that connect, sense and interact with the real world. Our ubiquitous IP portfolio and collaborative licensing model position us to capture secular growth across consumer IoT, automotive, industrial and infrastructure, and mobile and PC markets.

Our customers are increasingly receptive to our roadmap because it aligns with their need to add co

[Excerpt truncated for page length; the complete text is on the linked full-MD&A page.]

Read the full FY 2025 MD&A: /company/CEVA/mda/fy2025/
All MD&A years: /company/CEVA/mda/


## MD&A history

Prior-year 10-K MD&A spans are extracted from SEC filings with the same bounded parser used for the latest filing. Each year's full verbatim text is on its own sub-page.

- [FY 2024 MD&A](/company/CEVA/mda/fy2024/): filed 2025-02-27; accession 0001437749-25-005523 (https://www.sec.gov/Archives/edgar/data/1173489/000143774925005523/ceva20241231_10k.htm)
- [FY 2023 MD&A](/company/CEVA/mda/fy2023/): filed 2024-03-07; accession 0001437749-24-006970 (https://www.sec.gov/Archives/edgar/data/1173489/000143774924006970/ceva20231231_10k.htm)
- [FY 2022 MD&A](/company/CEVA/mda/fy2022/): filed 2023-03-01; accession 0001437749-23-005106 (https://www.sec.gov/Archives/edgar/data/1173489/000143774923005106/ceva20221231_10k.htm)
- [FY 2021 MD&A](/company/CEVA/mda/fy2021/): filed 2022-03-01; accession 0001437749-22-004905 (https://www.sec.gov/Archives/edgar/data/1173489/000143774922004905/ceva20211231_10k.htm)




## Macro cross-references

Indicators mapped to this company's SIC classification (industry 7370 Services-Computer Programming, Data Processing, Etc.) by grepcent's deterministic macro-sector crosswalk. A navigational mapping, not a statistical or causal claim.

- [PAYEMS](/indicator/PAYEMS/): All Employees, Total Nonfarm
- [CES0500000003](/indicator/CES0500000003/): Average Hourly Earnings of All Employees, Total Private
- [DGS10](/indicator/DGS10/): Market Yield on U.S. Treasury Securities at 10-Year Constant Maturity

Macro-to-micro threads including this sector: [US labor market](/thread/us-labor-market/), [Growth & output](/thread/growth-output/), [Government finances](/thread/government-finances/), [Sector employment](/thread/sector-employment/).

All macro indicators: /indicators/


## For LLMs & downloads

Markdown twin: /company/CEVA.md · JSON record: /company/CEVA.json · verified financials: /company/CEVA/financials.json / /company/CEVA/financials.csv · machine TOC for the whole site: /llms.txt