CEVA INC (CEVA) Business
This page reproduces the company's own Item 1 Business text from the linked SEC filing. It is filer text, not grepcent analysis, scoring, or investment advice.
Informational only - not investment advice. See Disclaimer.
ITEM 1. BUSINESS
Company Overview
We are the leader in silicon and software intellectual property (“IP”) enabling Physical AI - the intelligence embedded directly into billions of devices that connect, sense, and infer data in the real world. Physical AI represents the natural evolution of Edge AI. While traditional Edge AI focuses on local inference on devices, Physical AI unifies connectivity, sensing, and inference into a cohesive technology fabric. This integrated approach allows devices not only to process information at the edge but also to understand their physical surroundings, communicate seamlessly, and interact intelligently with both the environment and the cloud. As Physical AI adoption accelerates across markets, these foundational layers are becoming essential for next‑generation consumer, automotive, industrial, and mobile systems.
We are the only IP company with established leadership and deep expertise across all three foundational layers of Physical AI – connect, sense and infer. Our comprehensive portfolio powers the connectivity, perception, and intelligence capabilities of many of today’s most advanced smart edge products across the consumer IoT, automotive, industrial and infrastructure, and mobile and PC markets. Devices incorporating Ceva IP have shipped in more than 20 billion units since 2003, and today our technologies are embedded in approximately 2 billion devices annually, reflecting the ubiquity and long‑term durability of our IP across multiple generations of silicon and end‑market product cycles.
Our business model centers on licensing our IP platforms, processors, software, and related technologies to semiconductor and original equipment manufacturer (“OEM”) customers. These customers integrate our technologies into their application specific integrated circuits (“ASICs”), application specific standard products (“ASSPs”), and systems on chip (“SoCs”), which they manufacture, market, and sell globally. By providing pre‑validated, production‑proven IP and comprehensive software stacks, we enable our customers to reduce development time, improve performance and power efficiency, streamline certification and compliance, and bring differentiated products to market more reliably and cost‑effectively.
We operate with a global development footprint of more than 400 R&D, support, and sales professionals, reflecting the multidisciplinary expertise required to support wireless communications, advanced AI DSP architectures, AI inference engines, sensor fusion algorithms, and embedded software. Our headquarters is located in Rockville, Maryland, and we maintain development centers in France, Greece, Ireland, Israel, Serbia, and the United States. This globally distributed engineering model allows us to attract specialized talent, collaborate closely with customers in key regions, and support complex, multi‑site development programs.
We are a sustainability‑focused and environmentally conscious company. We have adopted a Code of Business Conduct and Ethics and a Sustainability Policy that emphasize environmental preservation, responsible material usage and recycling, employee welfare, data privacy, and ethical business practices. These principles guide our operations and culture at every level. We recognize that as an IP provider to high‑volume markets, our impact extends beyond our own operations and into the broader ecosystem; therefore, we remain committed to promoting responsible innovation and strengthening long‑term environmental and social stewardship.
Ceva’s corporate roots date back to its incorporation in Delaware on November 22, 1999, under the name DSP Cores, Inc. Our current company was formed in November 2002 through the combination of the DSP IP licensing division of DSP Group, Inc. and Parthus Technologies plc (“Parthus”). This merger created one of the industry’s first dedicated, pure‑play IP licensing companies, laying the foundation for Ceva’s growth into a global leader in the technologies powering Physical AI across the connect, sense, and infer layers.
3
Table of Contents
Industry Background
The proliferation of smart, connected devices across virtually every end market has established three essential capability layers that form the foundation of modern intelligent systems: Connect, Sense, and Infer. These layers define how devices communicate, perceive their environment, and generate real‑time intelligence at the edge.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| 1. | Connect - reliable, secure, high‑performance wireless connectivity Smart devices increasingly require multi‑standard wireless communications, spanning Wi‑Fi, Bluetooth, UWB, NB‑IoT, LTE Cat‑1, 5G, and 5G‑Advanced, each evolving rapidly and becoming more complex to implement. The growing range of applications - from consumer IoT devices to automotive systems and industrial automation - demands robust, low‑power connectivity solutions integrated directly into SoCs rather than through discrete chips. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| 2. | Sense - rich perception from audio, imaging, inertial, and RF sensors The rise of sensor‑rich products such as wearables, XR devices, smart cameras, robotics, and advanced automotive systems is driving demand for sophisticated sensor‑fusion algorithms, advanced audio processing, spatial awareness, and low‑latency perception pipelines. These workloads require specialized architectures optimized for combining data streams from cameras, microphones, inertial measurement units (IMUs), radar, and other sensors. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| 3. | Infer - real‑time, efficient on‑device intelligence through neural network and DSP processing As Edge AI workloads shift toward increasingly complex models - including generative, transformer‑based, and multi‑modal networks - devices require highly efficient NPUs and AI‑optimized DSPs capable of delivering significant computational throughput with ultra‑low power consumption. These capabilities underpin the industry’s broader transition from traditional Edge AI to Physical AI, where intelligence becomes an integral part of how devices interact with the real world. |
Across these layers, technological complexity continues to rise, driven by rapidly evolving global standards such as 5G/5G‑Advanced, Wi‑Fi 7/8, Bluetooth 6/7, UWB, and the growing sophistication of embedded AI/ML workloads. At the same time, semiconductor companies face increasing pressure to accelerate time‑to‑market, reduce engineering cost, and support expanding feature sets. This combination is widening the industry’s longstanding “design gap,” where the pace of semiconductor manufacturing advances outstrips the availability of specialized design resources required to fully harness them. As a result, device makers are turning to licensable, production‑proven IP to overcome development bottlenecks, reduce risk, and focus internal resources on system differentiation.
In parallel, the market opportunity continues to expand. The global Physical AI and Edge AI markets together are expected to reach approximately $170 billion by 2030, reflecting explosive growth in intelligent device categories. This expansion is being propelled by sustained advancements in mobile computing, industrial automation, automotive sensing, robotics, and the rapid adoption of both generative AI and embedded AI across a wide range of end markets.
Solutions Portfolio Overview
Our technology portfolio spans the full Physical AI stack and is organized into three solution pillars that collectively enable devices to connect, sense, and infer with high performance, low power, and exceptional reliability. These pillars reflect the company’s long‑standing leadership across wireless connectivity, sensing, and edge AI acceleration, and they support billions of devices annually across consumer IoT, mobile, PC, industrial, infrastructure, and automotive markets.
| Column 1 | Column 2 |
|---|---|
| 1. | Connect - Wireless Connectivity IP |
We believe we provide the industry’s broadest and most widely deployed portfolio of wireless connectivity and communications IP across short‑range, cellular IoT, and advanced 5G/5G‑Advanced domains. Our connectivity technologies are foundational to the smart edge ecosystem, powering mass‑market consumer audio devices, wearables, smartphones, IoT sensors, home networking equipment, industrial automation systems, and next‑generation infrastructure.
We hold the #1 worldwide market share in wireless connectivity IP in 2024 with 68% global share, reflecting unmatched scale, multi‑protocol expertise, and an exceptional track record of trust and reliability. Additional industry‑leading positions include:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | ~30% of worldwide Bluetooth IoT devices are powered by us; |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | ~50% of all True Wireless Stereo (TWS) earbuds (non‑Apple) integrate our Bluetooth and audio IP; |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | ~40% of worldwide cellular IoT devices are built on our modem and DSP platforms; and |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | ~30% of global 5G networks use our DSP technology for RAN baseband processing. |
4
Table of Contents
This scale creates a powerful flywheel: broad ecosystem familiarity, deep Tier‑1 trust, and faster adoption of new standards including Bluetooth 6/7, Wi‑Fi 7/8, UWB, 5G‑Advanced, and NTN (non‑terrestrial networks).
Key Connectivity Offerings
• Bluetooth Connectivity - Ceva‑Waves Bluetooth
Comprehensive hardware IP and software stacks supporting the latest Bluetooth and 802.15.4 standards, including high‑data‑throughput and LE Audio profiles used widely in TWS, wearables, toys, and smart home devices.
• Wi‑Fi Connectivity - Ceva‑Waves Wi‑Fi
Scalable subsystems supporting Wi‑Fi 4 through Wi‑Fi 7, optimized for both high‑performance applications (routers, access points, PCs) and low‑power embedded designs (IoT nodes, consumer electronics).
• Ultra‑Wideband - Ceva‑Waves UWB
MAC/PHY IP enabling precise ranging, secure localization, device‑finding, indoor navigation, and Doppler‑based presence detection for next‑generation consumer, safety, and industrial applications.
• Multi‑Protocol Connectivity - Ceva‑Waves Links
Integrated multi‑radio platforms with advanced coexistence algorithms allowing Bluetooth, Wi‑Fi, UWB, and other radios to operate reliably on shared RF front‑ends - ideal for wearables, TWS earbuds, and IoT devices.
• Cellular IoT - Ceva‑Waves DragonFly
Turnkey NB‑IoT modem platforms with GNSS support and integrated sensor‑fusion capabilities, powering large deployments across smart metering, asset tracking, environmental sensing, and industrial IoT.
• 5G User Equipment & Infrastructure
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Ceva‑XC20 DSPs and the PentaG2 platform support 5G/5G‑Advanced baseband for mobile broadband, RedCap, and IoT devices. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | PentaG‑RAN delivers carrier‑grade Layer‑1 PHY processing for small cells, Open RAN, private networks, and satellite communications. |
These platforms dramatically reduce design risk, accelerate time‑to‑market, and provide production‑proven silicon‑ready IP for companies entering or scaling in advanced communications markets.
| Column 1 | Column 2 |
|---|---|
| 2. | Sense - Sensing, Perception & Audio Processing |
We provide a comprehensive suite of DSPs and embedded software solutions that transform raw sensor data into actionable intelligence. Our technologies support key perception workloads - including vision, radar, IMU fusion, audio enhancement, spatial computing, and environmental sensing - used across wearables, smartphones, XR devices, robotics, industrial systems, and automotive ADAS.
With over 500 million devices powered to date by our sensing and audio embedded software and hundreds of millions of devices powered by our AI DSPs, the company has established itself as a global leader in motion tracking, spatial audio, and multimodal sensor processing.
Key Sensing Offerings
• Ceva‑SensPro AI and Sensor Fusion DSPs
High‑performance DSP cores that combine vision processing, radar signal processing, and AI workloads in a unified architecture - ideal for automotive ADAS, robotics, smart cameras, and industrial inspection.
• Ceva‑MotionEngine
High‑accuracy IMU sensor‑fusion software delivering precise 6‑ and 9‑axis motion tracking for wearables, controllers, XR devices, drones, and smart home interfaces.
• Ceva‑RealSpace
Advanced spatial audio and head‑tracking software providing immersive, low‑latency 3D audio experiences for AR/VR, TWS earbuds, gaming headsets, and home entertainment systems.
5
Table of Contents
• Ceva‑BX1/BX2 Audio & Control AI DSPs
Low‑power audio and control processors enabling voice activation, neural‑network ENC, audio pre/post‑processing, and efficient microcontroller‑class workloads.
Collectively, these technologies enable rich multimodal perception and contextual awareness in mobile devices, automobiles, industrial robots, smart cameras, and factory automation systems.
3. Infer - AI Processing & Neural Network Acceleration
Our inference portfolio addresses the full AI performance spectrum - from ultra‑efficient embedded AI workloads to high‑performance generative AI and automotive‑grade inferencing.
Supported by strong adoption momentum and marquee licensees, our NPU roadmap positions the company as a foundational engine for device‑side AI acceleration.
Key AI Inference Offerings
• NeuPro‑Nano
An ultra‑efficient embedded NPU architecture optimized for AIoT workloads across audio, vision, sensor analytics, anomaly detection, and predictive maintenance. Ideal for microcontrollers, TWS earbuds, wearables, smart home, and edge sensors.
• NeuPro‑M
A scalable, high‑performance NPU for generative AI, transformers, Vision Transformers (ViT), multimodal networks, and automotive‑grade applications including ADAS and in‑cabin intelligence. Tailored for PCs, smartphones, gateways, robotics, and advanced industrial platforms.
• NeuPro Studio SDK
A unified toolchain enabling training, optimization, quantization, graph compilation, simulation, and deployment across all NeuPro NPUs. Simplifies AI model integration and shortens customers’ time‑to‑market.
These inferencing solutions allow customers to run advanced AI models on devices with:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Lower latency |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Higher privacy and security |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Reduced cloud dependence and cost |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Optimized power efficiency |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Scalable performance from milliwatts to watts |
Markets and Applications
Our based solutions serve diversified, high‑volume markets:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Consumer IoT - hearables, wearables, smart home devices, TWS earbuds, cameras, XR products |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Automotive - ADAS, infotainment, sensor fusion, V2X, vehicle connectivity |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Industrial & IIoT - automation, robotics, predictive maintenance, smart factories |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Infrastructure - 5G RAN, fixed wireless access, satellite communications, private networks |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Mobile & PC - smartphones, tablets, AI PCs, modems, connectivity subsystems |
Growing device intelligence, coupled with Physical AI adoption, continues to expand our customer pipeline and addressable market across all these categories.
Business Model
We operate under a licensing and royalty model:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Licensing & related revenue - Upfront fees for access to our platforms, IP cores, software, and development tools, including customization services and technical support. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Royalty revenue - Recurring per‑unit or percentage‑based royalties on customer shipments incorporating our IP. |
6
Table of Contents
This model enables:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Scalable, long‑term revenue |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Technology reuse across markets |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Broad customer reach without manufacturing risk |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | A sustainable R&D‑driven growth engine |
Strategy
Our strategy is centered on capturing the accelerating, secular demand for intelligent, connected devices by leveraging our unique leadership position across the full Physical AI technology stack. As end markets transition from basic connected devices to context‑aware, AI‑enabled smart edge systems, our strategy focuses on expanding our influence across the three foundational layers - connect, sense, and infer - while deepening our role as a critical technology partner to global semiconductor and OEM ecosystems.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Extending leadership across connect, sense, and infer technologies |
We are the #1 worldwide provider of wireless connectivity IP, spanning Bluetooth, Wi‑Fi, UWB, cellular IoT, and 5G RAN DSP platforms, positioning the company as the industry’s de‑facto connectivity engine for the smart edge. Building on this leadership, we continue to enhance our connectivity, sensing, and inference platforms to address rising requirements in multi‑protocol coexistence, spatial awareness, and real‑time on‑device intelligence.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Scaling adoption of NeuPro and SensPro AI families to address both embedded and generative AI at the edge |
With AI adoption expanding from premium devices to mass‑market microcontrollers, IoT sensors, and automotive systems, we are accelerating deployment of NeuPro‑Nano for embedded ML and NeuPro‑M for high‑performance generative AI, ADAS, and multimodal workloads.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Driving broader deployment of Wi‑Fi 6/7, Bluetooth, UWB, and 5G‑Advanced connectivity |
As the global transition to next‑generation connectivity standards accelerates, we are expanding our roadmap and customer engagements across Wi‑Fi 6/7, Bluetooth 6/7, UWB, 5G‑Advanced, and non‑terrestrial networks. We are positioned to capture long‑term royalty uplift as device volumes scale and customers adopt more advanced radios.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Expanding platform‑level solutions that combine processors, PHY/MAC components, RF, software, tools, and reference designs |
Customers increasingly demand integrated, silicon‑ready subsystems that combine compute engines, connectivity IP, software stacks, RF components, and tools. We continue to expand full‑stack platforms (e.g., Ceva‑Waves Bluetooth/Wi‑Fi/UWB, PentaG2, SensPro, NeuPro Studio) that reduce development time, complexity, and certification overhead while increasing Ceva’s value capture per design.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Increasing software monetization across audio, spatial computing, and sensor fusion |
With OEMs’ requirements for advanced embedded application software to differentiate their devices through user experiences, we are expanding monetization of software assets - including RealSpace, MotionEngine, and NeuPro Studio - to enable richer audio, perception, and AI experiences. These software layers strengthen customer lock‑in and provide recurring, differentiated value independent of silicon cycles.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Deepening relationships across the world’s largest MCU, SoC, and consumer‑device ecosystems |
Our IP is already adopted by seven of the world’s top 10 MCU vendors, as well as global PC, smartphone, wearables, and IoT OEMs. Our strategy is to broaden our presence across these ecosystems, supporting cross‑portfolio adoption of connectivity IP, AI DSPs, and NPUs to increase content per device and expand multi‑technology engagements.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Targeting new growth verticals such as robotics, spatial computing, next‑generation PCs, and industrial automation |
Emerging device categories - including autonomous robotics, immersive XR, AI PCs, advanced wearables, and Industry 4.0 systems - require tightly integrated connectivity, sensing, and inference. Our Physical AI Fabric and multi‑domain expertise position the company to become a foundational technology supplier in these fast‑growing markets.
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Maintaining disciplined investment in high‑leverage technologies aligned to Physical AI trends |
We continue to invest in scalable, high‑impact technologies that enable the next era of edge intelligence, including multi‑protocol RF architectures, generative AI NPUs, sensor‑rich perception engines, and standards‑driven wireless platforms. These investments reinforce our long‑term position as the leading provider of IP enabling Physical AI across markets.
7
Table of Contents
Customers
We have licensed our IP to hundreds of semiconductor and OEM customers worldwide, including seven of the world’s top ten MCU vendors, with adoption spanning mobile, audio, consumer IoT, wearables, industrial automation, automotive, and infrastructure markets. Our publicly announced licensees include companies such as Actions, Ambiq, ASPEED, ASR Microelectronics, Atmosic, Beken, Bestechnic, boAt, Broadcom, Ceragon, Cirrus Logic, Espressif Systems, ESWIN, Fujifilm, Rockchip, Goodix, Hisense, iCatch Technology, iCOM, InPlay, Intel, Itron, LG Electronics, LifeSignals, ListenAI, MediaTek, Microchip, Nations Technologies, Nextchip, Nordic Semiconductor, Nothing, Novatek, NXP, onsemi, Oticon, Panasonic, Qualcomm, Realtek, Renesas, Rohm, Samsung, SatixFy, Sharp, SiFli, SigmaStar, Socionext, Sony, Sonova, STMicroelectronics, Synaptics, THX, Toshiba, Unisoc, Vatics, Winner Micro, and Yamaha. This diverse and global set of partners highlights our deep entrenchment across the semiconductor value chain and our critical role in enabling Physical AI capabilities in billions of smart edge devices.
International Sales and Operations
Customers based in Europe and the Middle East (EME) and Asia Pacific (APAC) accounted for 82% of our total revenues for 2025 and 81% of our total revenues for 2024, with customers in China accounting for 62% and 49% of total revenues for 2025 and 2024, respectively. Additional information on the geographic breakdown of our revenues and location of our long-lived assets is contained in Note 12 to our consolidated financial statements, which appear elsewhere in this annual report.
8
Table of Contents
Sales and Marketing
We license our technology through a direct sales force. As of December 31, 2025, we had 30 employees in sales and marketing. We have sales offices and representation in the APAC region, Sweden, France, Israel, the United Kingdom and the United States.
Maintaining close relationships with our customers and strengthening these relationships are central to our strategy. From time to time, we develop new signal processors, AI solutions, platforms, software solutions or connectivity products in close alignment with several tier-one industry players, which signifies to the market that we are focused on viable applications that meet broad industry needs or try to get similar inputs and insight for our new developments from our marketing team. Generally, these industry leaders become licensees for these products which allows us to create a roadmap for the future development of different IP solutions and helps us to anticipate the next potential applications for the market. We seek to use our customer relationships to deliver new products in a faster time to market.
We use a variety of marketing initiatives to stimulate demand and brand awareness in our target markets. These marketing efforts include contacts with industry analysts, presenting at key industry trade shows and conferences, and a comprehensive digital marketing program aimed at developing and nurturing relationships with potential customers. Our marketing group runs competitive benchmark analyses to help us maintain our competitive position.
Technical Support
We offer technical support services through our offices in Israel, the APAC region, France and the United States. As of December 31, 2025, we had 32 employees in technical support. Our technical support services include:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | assistance with implementation, responding to customer-specific inquiries, training and, when and if they become available, distributing updates and upgrades of our products; |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | application support, consisting of providing general hardware and software design examples, ready-to-use software modules and guidelines to our licensees to assist them in using our technology; and |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | design services, consisting of creating customer-specific implementations of our signal processing AI IP and application platforms. |
We believe that our technical support services are a means to assist our licensees to embed our IP solutions in their designs and products. Our technology is highly complex, combining sophisticated signal processing IP core architectures, AI solutions, integrated circuit designs and development tools. Effective customer support in helping our customers to implement our solutions enables them to shorten the time to market for their applications. Our support organization is made up of experienced engineers and professional support personnel. We conduct technical training for our licensees and their customers and meet with them from time to time to track the implementation of our technology.
Research and Development
R&D is the core of our value proposition. Our innovation engine is driven by a highly specialized global engineering organization, comprising more than 300 dedicated R&D professionals with deep expertise in wireless communications, signal processing, AI compute, sensing, and embedded software. We also supplement our internal R&D organization with specialized external contractors as required. The skill sets required to design world‑class wireless IP, domain‑specific processors, and advanced AI architectures are scarce and increasingly difficult for semiconductor companies to build and retain internally. Our concentrated talent pool - spanning algorithm development, PHY/MAC and RF design, DSP and NPU architecture, toolchain development, and system integration - allows us to deliver production‑ready, standards‑compliant IP that is exceptionally hard for customers to replicate in‑house.
Our engineering teams focus on:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Wireless communications standards, including Bluetooth, Wi‑Fi, UWB, cellular IoT, and 5G/5G‑Advanced |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | AI DSP architectures for audio, vision, radar, and multimodal sensing |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | AI processors and toolchains, including embedded ML, classical AI, and generative AI NPUs |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Sensor‑fusion engines for motion, environmental, and spatial computing applications |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Audio and spatial computing technologies, enabling immersive 3D audio, voice, and advanced noise processing |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Embedded software, development environments, and SDKs, including AI compilers, debug tools, and reference frameworks |
9
Table of Contents
Because these domains require deep specialization and long development cycles, customers increasingly rely on us to supply the underlying connectivity, sensing, and inference technologies rather than dedicating internal R&D resources to building and maintaining these capabilities themselves. By licensing our proven, silicon‑ready IP platforms, customers can accelerate time‑to‑market, reduce risk, and redeploy their engineering investments toward differentiated product features such as UI/UX, system‑level optimization, proprietary applications, and platform integration.
In addition to internal innovation, we participate actively in global standards bodies and industry consortiums. Our involvement ensures early visibility into emerging specifications, enables us to influence future standards, and helps maintain our leadership position as new wireless, sensing, and AI capabilities evolve.
Our R&D organization remains one of our most important strategic assets - enabling us to deliver high‑performance, low‑power IP that shifts the economics of semiconductor design and empowers our customers to innovate more efficiently and competitively.
Competition
The markets in which we operate are intensely competitive. They are subject to rapid change and are significantly affected by new product introductions. We compete with other suppliers of licensed signal processing IP, licensed processing AI and wireless connectivity IP. We believe that the principal competitive elements in our field are signal processing IP and NPU performance, overall chip cost, power consumption, flexibility, reliability, communication and multimedia software and algorithms availability, design cycle time, tool chain, customer support, financial strength, name recognition and reputation. We believe that we compete effectively in each of these areas but can offer no assurance that we will have the financial resources, technical expertise, and marketing or support capabilities to compete successfully in the future.
The markets in which we compete are dominated by large, highly competent semiconductor companies that have significant brand recognition, a large installed base and a large network of support and field application engineers. The following industry players and factors may have a significant impact on our competitiveness:
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | DSP, NPU, and processor IP vendors, including Arm, Cadence, Imagination Technologies, VeriSilicon, Andes Technology, SiFive, Expedera, Quadric and Digital Media Professionals (DMP), as well as GlobalFoundries, which acquired certain Synopsys IP businesses and now participates more directly in the semiconductor IP ecosystem. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | AI accelerator providers, including Arm Ethos, Cadence AI/Vision DSP families, Imagination AI accelerators, Cambricon, Ambarella (CVflow), and open‑source engines such as Nvidia’s NVDLA. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Short‑range wireless IP suppliers such as Mindtree, along with internal wireless design teams at major semiconductor companies who develop Bluetooth, Wi‑Fi, UWB, and multi‑protocol solutions in‑house. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Companies offering proprietary DSP, AI, or connectivity architectures developed internally, especially large integrated semiconductor manufacturers that design their own radios, DSPs, sensor‑fusion engines, audio pipelines, or NPUs. |
| Column 1 | Column 2 | Column 3 |
|---|---|---|
| ● | Software providers in spatial audio, sensing, and motion‑processing, including Dolby, Waves, Dirac, and CyweeMotion, as well as internal algorithm teams within leading OEMs. |
Competitive differentiation is based on performance, power efficiency, flexibility, time‑to‑market, standards leadership, software ecosystem depth, developer tools, and quality of customer support.
In addition, we may face increased competition from smaller, niche semiconductor design companies in the future. Some of our customers also may decide to satisfy their needs through in-house design. We compete based on signal processing IP performance, first-to-market availability for latest generation wireless standards, overall chip cost, power consumption, flexibility, reliability, communication and multimedia software availability, design cycle time, tool chain, customer support, name recognition, reputation and financial strength. Our inability to compete effectively on these bases could have a material adverse effect on our business, results of operations and financial condition.
With respect to certain large potential customers, we also compete with internal engineering teams, which may design wireless connectivity or programmable signal processing IP core products in-house. Companies such as NXP and STMicroelectronics license our designs for some applications and use their own proprietary cores for other applications. These companies also may choose to license their wireless connectivity or proprietary signal processing IP cores to third parties and, as a result, become direct competitors.
10
Table of Contents
Proprietary Rights
Our success and ability to compete are dependent on our ability to develop and maintain the proprietary aspects of our technology and to operate without infringing the proprietary rights of others. We rely on a combination of patent, trademark, trade secret and copyright laws and contractual restrictions to protect the proprietary aspects of our technology. These legal protections afford only limited protection of our technology. We also seek to limit disclosure of our IP and trade secrets by requiring employees and consultants with access to our proprietary information to execute confidentiality agreements with us and by restricting access to our source code and other IP. Due to rapid technological change, we believe that factors such as the technological and creative skills of our personnel, new product developments and enhancements to existing products are more important than obtaining statutory protections for our technology in establishing and maintaining a technology leadership position.
We have an active program to protect our proprietary technology through the filing of patents. Our patents relate to our signal processing IP cores, AI solutions and application-specific platform technologies. As of December 31, 2025, we held 48 patents in the United States, nine patents in Canada, 98 patents in the EME region and 16 patents in the Asia Pacific (APAC) region, totaling 171 patents, with expiration dates between 2026 and 2041. In addition, as of December 31, 2025, we had six patent applications pending in the United States, five pending patent applications in the EME region, two pending global Patent Corporation Treaty (PCT) patent applications and two pending patent applications in the APAC region, totaling 15 pending patent applications.
We actively pursue foreign patent protection in countries where we feel it is prudent to do so. Our policy is to apply for patents or for other appropriate statutory protection when we develop valuable new or improved technology that is eligible for protection. The patent prosecution process involves complex legal and factual questions, and the breadth of claims allowed is uncertain. Accordingly, there are no assurances that any patent application filed by us will result in a patent being issued, or that our issued patents, and any patents that may be issued in the future, will afford us adequate protection against competitors with similar technology; nor can we be assured that patents issued to us will not be infringed or invalidated, or that others will not design around our technology. In addition, the laws of certain countries in which our products are or may be developed, manufactured or sold may not protect our products and IP rights to the same extent as the laws of the United States. We can provide no assurance that our pending patent applications or any future applications will be approved or will not be challenged by third parties, that any issued patents will effectively protect our technology, or that patents held by third parties will not have an adverse effect on our ability to do business.
The semiconductor industry is characterized by frequent litigation regarding patent and other IP rights. Questions of infringement in the semiconductor field involve highly technical and subjective analyses. In addition, patent infringement claims are increasingly being asserted by patent holding companies (so-called patent “trolls”), which do not use technology and whose sole business is to enforce patents against companies, such as us, for monetary gain. Because such patent holding companies do not provide services or use technology, the assertion of our own patents by way of counterclaim may be ineffective. Litigation may in the future be necessary to enforce our patents and other IP rights, to protect our trade secrets, to determine the validity and scope of the proprietary rights of others, or to defend against claims of infringement or invalidity. We cannot assure you that we would be able to prevail in any such litigation.
In any potential dispute involving our patents or other IP, our licensees also could become the targets of litigation. We are generally obligated to indemnify our licensees under the terms of our license agreements. Although our indemnification obligations are typically subject to a maximum amount, these obligations could nevertheless result in substantial expenses. In addition to the time and expense required for us to indemnify our licensees, a licensee’s development, marketing and sale of products embodying our solutions could be severely disrupted or shut down as a result of litigation, and we could be liable to such licensees for additional amounts.
We also rely on trademark, copyright and trade secret laws to protect our IP. We have registered trademarks in the United States for our name Ceva and the related Ceva logo and currently market our signal processing cores and other technology offerings under this trademark.
11
Table of Contents
Human Capital Resources
The table below presents the numbers of our employees as of December 31, 2025 by function and geographic location.
| Number | |||
|---|---|---|---|
| Total employees | 424 | ||
| Function | |||
| Research and development | 311 | ||
| Sales and marketing | 30 | ||
| Administration | 51 | ||
| Technical support | 32 | ||
| Location | |||
| Europe, Middle East (EME) | 363 | ||
| Asia Pacific (APAC) | 33 | ||
| United States | 28 |
We believe we are a respected employer in the countries where we have operations, and, with the help of our employees, we strive to be a responsible global corporate citizen and a more sustainable company. Our Code of Business Conduct and Ethics sets the standards of conduct of our directors, officers and employees. In addition, we adopted a Sustainability Policy that addresses matters related to our employees as well as data privacy and security, resource conservation and recycling, and other environmental matters. Our Sustainability Policy reflects our commitment to diversity and equal opportunity, a harassment-free workplace, training, development and employee engagement, and human rights, health and safety, and other matters relevant to employee well-being and the Ceva culture. The code is reviewed and updated periodically by our Board or Directors, and both the code and our Sustainability Policy are available on our website at www.ceva-ip.com.
Our employees are not represented by any collective bargaining agreements, however, certain provisions of Israeli law and the collective bargaining agreements between the Histadrut (General Federation of Labor in Israel) and the Coordination Bureau of Economic Organizations (the Israeli federation of employers’ organizations) apply to our Israeli employees. We have never experienced a work stoppage. We believe our employee relations are good, as is their general well-being, which is one of management’s top priorities.
Available Information
Our annual reports on Form 10-K, quarterly reports on Form 10-Q, current reports on Form 8-K and amendments to reports pursuant to Section 13(a) or 15(d) of the Securities Exchange Act of 1934, as amended, are available, free of charge, on our website at www.ceva-ip.com, as soon as reasonably practicable after such reports are electronically filed with the Securities and Exchange Commission and are also available on the SEC’s website at www.sec.gov.
Our website and the information contained therein or connected thereto are not intended to be incorporated into this Annual Report on Form 10-K.