12:30 – 7:30 pm (Technology Exhibits Reception 6 – 7:30 pm)
11 am – 5 pm
AMD is redefining intelligent edge and physical AI by bringing data center-class performance to embedded platforms without compromise. We will demonstrate how our ROCm open-source stack unlocks the full compute potential of Radeon iGPUs for real-time AI/ML on power-constrained hardware. Live demos will showcase vision and LLM workloads running locally on Ryzen AI, highlighting an open, flexible stack for deploying any model topology from vision to LLMs at the edge.
Booth: #516
Your trusted source for analysis, advice and engineering for physical AI and computer vision. For more than 30 years we’ve helped companies develop, choose and use embedded technology. Our expertise in architectures, algorithms, models and tools helps clients reach the market faster, with less risk and better products. We’ve helped clients build vision systems from driver safety to kitchen appliances, on platforms from microcontrollers to FPGAs to servers. What can we do for you?
Booth: #521
Boston.AI delivers advanced machine vision and edge AI for embedded systems. Experience machine vision combining LiDAR and camera data for precise spatial awareness, sensor fusion and interactive insights on-device without cloud dependency. Explore our AI photo kiosk, transforming live headshots into AI-generated art with real-time image processing and a touchscreen interface.
Booth: #417
Cadence Tensilica is a global leader in configurable digital signal processor (DSP) IP. Based on the versatile Xtensa architecture, developers can generate customized processors for AI, audio, computer vision, radar and many more applications. Key offerings include the HiFi DSP for audio, voice and speech; the Vision DSP for imaging and radar; and the AI Platform for on-device AI processing.
Booth: #402
Efinix is an innovator in efficient FPGA technology. With three product families ranging from 4K logic elements (LEs) up to 1M LEs, our FPGAs span the entire market from entry level to high-end applications. The flagship Titanium family features products with hardened quad-core RISC-V processors, LPDDR4x controllers and high-speed SerDes. The free downloadable tools provide an intuitive and high-performance path from HDL to bitstream with integrated C/C++ flows for the RISC-V and turnkey AI acceleration.
Booth: #503
Discover our industry-leading low-power, small form factor FPGAs and solutions for automotive, industrial and security applications at the edge. Come see our technologies in action—from cloud to sensor, including low-power FPGAs, solution stacks, software tools, partner solutions and more.
Booth: #416
MagikEye is a 3D sensing company developing ILT (Invertible Light Technology), a depth perception approach that shifts computation from digital processing into the optical domain. By combining encoded illumination with standard CMOS image sensors, ILT enables real-time depth extraction with reduced compute, power, system cost and memory requirements. Unlike stereo, structured light or time-of-flight, ILT minimizes system complexity, enabling scalable deployment of efficient 3D vision across robotics, automotive, consumer and physical AI platforms.
Booth: #316
Smarter embedded vision begins with the right foundation. We’ll be showcasing how our power-efficient FPGAs and secure interfaces enable real-time vision across edge AI, automotive and industrial applications. Visit our booth to experience face recognition with liveness detection, high-speed SLVS-EC and CoaXPress vision bridges and low-latency sensor pipelines for edge AI.
Booth: #302
Micron will showcase how advanced memory and storage solutions enable the next generation of edge AI, computer vision and intelligent robotics. At our booth, we’ll highlight the critical role of high‑performance, power‑efficient memory in meeting the growing demands of vision‑based AI workloads running at the edge.
Booth: #303
Experience our cutting-edge lineup featuring solid-state LiDAR for immersive wide-angle imaging and high-performance drone AF modules for precision aerial focus. We will also present our integrated ToF + RGB sets, delivering superior depth sensing and color accuracy, alongside our secure face recognition solution. These innovations exemplify our commitment to providing high-quality, AI-integrated camera modules that empower smart connectivity and security across diverse industries.
Booth: #509
Nota AI is a leading innovator in AI optimization and compression. Its proprietary AI model optimization platform, NetsPresso, enables high-performance AI models to run efficiently on resource-constrained edge devices through automated optimization and deployment. Nota Vision Agent (NVA), a VLM-powered video monitoring solution, delivers real-time insights across industries, while VLA advancements extend AI from perception to real-world action toward physical AI.
Booth: #408
We built Avocado OS to make scaling computer vision easy—production-ready Linux with over-the-air (OTA) updates, secure boot and fleet management built in. Pre-validated CUDA, TensorRT and OpenCV across NVIDIA Jetson, NXP and Qualcomm mean your team skips BSP integration entirely. Avocado OS decouples the OS layer from your application logic so ML engineers iterate on models and inference pipelines, not kernel drivers and meta-layers. Go from EVK to production deployment in weeks, not months.
Booth: #609
Synetic provides world-class data and computer vision solutions that help organizations build, train and deploy AI systems for real-world applications. We support customers with the data, tools and technical expertise needed to improve vision-based AI performance across a range of industries.
Booth: #309
Synopsys ARC NPX6 NPU IP addresses the demands of real-time compute with ultra-low-power consumption for AI applications. To accelerate application software development for the ARC NPX6 NPU IP, the ARC MetaWare MX Development Toolkit provides a comprehensive compilation environment with automatic neural network algorithm partitioning to maximize resource utilization.
Booth: #616
VeriSilicon offers six in-house processing IPs—GPU, NPU, VPU, DSP, ISP and display—plus 1,700+ analog, mixed-signal, RF and interface IPs. Through our SiPaaS model, we serve diverse markets, including consumer electronics, automotive, computing, industrial, data processing and IoT, supporting customers such as fabless firms, IDMs, OEM/ODMs, internet companies and cloud providers.
Booth: #508
3LC empowers AI builders with seamless tools to debug, diagnose and improve training data directly inside their workflows. By leveraging model feedback during training, we radically reduce false positives, accelerate training up to 75% and deliver smaller, more accurate models—without data duplication or workflow disruption.
Booth: #517
Experience the future of physical AI with BrainChip’s Akida technology—neuromorphic efficiency that delivers always-on intelligence at the edge. Designed for battery-constrained devices, Akida enables continuous, real-time performance with zero wake-up delays, on-device sensor processing and data sovereignty built in. Sensitive information never reaches external servers.
Booth: #608
Qualitas Semiconductor delivers high-performance interface IP solutions, including SerDes, PHY and protocol controllers for PCIe, Ethernet, USB, MIPI and Display standards. Our IP emphasizes low-power, robust signal integrity, comprehensive verification and seamless SoC integration, enabling scalable connectivity across data center, AI, mobile and automotive applications.
Booth: #602
Squint Insights Studio is a software platform bringing explainability and oversight to AI applications. With our studio, developers iterate more quickly, identify where to improve their models and datasets, cut project costs by up to 40% and deploy pipelines automating error detection and correction, increasing overall trust and safety.
Booth: #617
Experience ST’s latest embedded vision innovations: our smart rearview camera with real-time object detection and counting for smarter, safer mobility; our latest time-of-flight (ToF) product, the VL53L9CX dToF 3D mini-LiDAR module, delivering high-resolution depth maps for robotics, AR and more; and our ultra-low-power smart doorbell featuring on-device edge AI for presence and parcel detection—all powered by the latest STM32N6 MCU.
Booth: #502
3LC empowers AI builders with seamless tools to debug, diagnose and improve training data directly inside their workflows. By leveraging model feedback during training, we radically reduce false positives, accelerate training up to 75% and deliver smaller, more accurate models—without data duplication or workflow disruption.
Booth: #517
Aion Silicon is a trusted partner in high-performance semiconductor design, specializing in advanced SoC solutions—including tailored ASICs—for AI, automotive, HPC, 5G, networking and other applications. Our full-service, high-touch engineering model with consultative project leadership guides customers from SoC architecture and IP selection through design, foundry tape-out and production.
Booth: #803
AIRY3D will demonstrate DepthIQ, our single-sensor, passive RGB + depth solution delivering high-quality 3D on any sensor. Come see demos showcasing DepthIQ running on camera, laptop, Qualcomm RB5 and NVIDIA Jetson platforms targeting consumer robotics, automotive, industrial and laptop markets with low-cost, compact and scalable 3D vision.
Booth: #420
Alif Semiconductor delivers next-generation AI-enabled 32-bit microcontrollers that unify high-performance compute, advanced security, connectivity and ultra-low-power efficiency in a single scalable architecture, empowering developers to build intelligent, always-on edge devices across consumer, industrial and enterprise applications.
Booth: #505
AMD is redefining intelligent edge and physical AI by bringing data center-class performance to embedded platforms without compromise. We will demonstrate how our ROCm open-source stack unlocks the full compute potential of Radeon iGPUs for real-time AI/ML on power-constrained hardware. Live demos will showcase vision and LLM workloads running locally on Ryzen AI, highlighting an open, flexible stack for deploying any model topology from vision to LLMs at the edge.
Booth: #516
Andes Technology, a Founding Premier member of RISC-V International, offers industry-leading RISC-V CPU IP—from ultraefficient 32-bit to high-performance 64-bit out-of-order multicore clusters featuring vector processing, AI stack, ACE customization and ISO 26262 compliance. Over 19 billion Andes-powered SoCs deployed globally.
Booth: #305
Stop by our booth, where we will be showcasing the Arducam qb family of all-in-one edge AI cameras and Arducam RoAD—an end-to-end vision platform for robotics—featuring our high-performance sensor-based camera collection (IMX500, ISX031, IMX900 and others), multi-camera systems, MCU cameras, UVC cameras, depth cameras and more.
Booth: #702
ASYGN will showcase ColibryNPU, which brings intelligence to electronic devices, allowing them to understand what they see with minimal energy. ColibryNPU is recognized as the most power-efficient AI chip for vision, bringing the next generation of devices to another level—from smart glasses and drones to robots and smart cameras.
Booth: #904
Your trusted source for analysis, advice and engineering for physical AI and computer vision. For more than 30 years we’ve helped companies develop, choose and use embedded technology. Our expertise in architectures, algorithms, models and tools helps clients reach the market faster, with less risk and better products. We’ve helped clients build vision systems from driver safety to kitchen appliances, on platforms from microcontrollers to FPGAs to servers. What can we do for you?
Booth: #521
Boston.AI delivers advanced machine vision and edge AI for embedded systems. Experience machine vision combining LiDAR and camera data for precise spatial awareness, sensor fusion and interactive insights on-device without cloud dependency. Explore our AI photo kiosk, transforming live headshots into AI-generated art with real-time image processing and a touchscreen interface.
Booth: #417
Experience the future of physical AI with BrainChip’s Akida technology—neuromorphic efficiency that delivers always-on intelligence at the edge. Designed for battery-constrained devices, Akida enables continuous, real-time performance with zero wake-up delays, on-device sensor processing and data sovereignty built in. Sensitive information never reaches external servers.
Booth: #608
Cadence Tensilica is a global leader in configurable digital signal processor (DSP) IP. Based on the versatile Xtensa architecture, developers can generate customized processors for AI, audio, computer vision, radar and many more applications. Key offerings include the HiFi DSP for audio, voice and speech; the Vision DSP for imaging and radar; and the AI Platform for on-device AI processing.
Booth: #402
Chips&Media is a premium multimedia IP leader specializing in HW video codecs, image processing NPUs and frame buffer compression (FBC) IPs since 2003. With 3B+ units powered globally, we provide high-resolution solutions (up to 8K) for 150+ partners. Our innovative custom NPUs now deliver high-quality imaging capabilities optimized for next-generation edge devices.
Booth: #605
Commonlands offers high-performance, cost-effective M12 lenses, C-mount lenses and camera modules. We provide commercial off-the-shelf optics with fast shipping from the US with domestic engineering support. We also offer domestic lens design and production. Our products are used in robotics, IoT, smart retail, factory automation, drones and more. Stop by our booth to learn more.
Booth: #808
SAFEflow is a vision-based infusion system that uses real-time observation of fluid behavior as the feedback signal for closed-loop flow control, bringing direct measurement and intelligent sensing to a clinical process that has historically relied on open-loop control. Visit our booth to learn more.
Booth: #619
Join us at our booth for the unveiling of our RIVER silicon technology for embedded vision processing in robotics and autonomous devices. Optical flow is a key component of many autonomous navigation algorithms, and we will be demonstrating RIVER applied to it using an end-to-end event-based pipeline, from sensing to processing.
Booth: #801
Efinix is an innovator in efficient FPGA technology. With three product families ranging from 4K logic elements (LEs) up to 1M LEs, our FPGAs span the entire market from entry level to high-end applications. The flagship Titanium family features products with hardened quad-core RISC-V processors, LPDDR4x controllers and high-speed SerDes. The free downloadable tools provide an intuitive and high-performance path from HDL to bitstream with integrated C/C++ flows for the RISC-V and turnkey AI acceleration.
Booth: #503
eInfochips, an Arrow company, provides solutions across the product stack—from edge to cloud. We will be showcasing our niche machine vision and edge AI solutions, developed leveraging 30+ years of engineering excellence, reference designs and framework based on the latest platforms.
Booth: #902
ENERZAi offers full-stack edge AI software technology spanning model design, training, quantization and compile-level optimization to deal with memory bottlenecks on the edge. Stop by our booth, where we will be showcasing our edge audio and language AI suite, including voice control, live translation, AI agent and more.
Booth: #518
Expedera provides customizable neural engine semiconductor IP that dramatically improves performance, power and latency while reducing cost and complexity. Expedera’s NPU solutions are scalable and produce superior results in edge nodes and smartphones to automotive applications.
Booth: #520
Turn footage and images into decision-ready insights so you can move faster, reduce risk and prove what happened (without building a vision team). Our Visual Intelligence API is designed to help teams scale visual insight responsibly and get consistent results, clearer reporting and workflows leaders can rely on.
Booth: #704
Originating from a lineage of advanced image and signal processing, FotoNation’s sixth-generation AI IP empowers low-power, low-latency and high-performance sensor compute. These IPs are meticulously crafted for the contemporary era of “companion and physical AI,” characterized by perpetual human-like perception. Stop by our booth for the unveiling.
Booth: #807
Grinn offers comprehensive services in hardware design, embedded software development, mechanical engineering and manufacturing support. We specialize in edge AI solutions for IoT, robotics, industrial systems, human-machine interfaces, computer vision, security and high-performance AI applications.
Booth: #806
Our test and calibration portfolio includes truly unique solutions for edge AI camera development. Stop by to see our lens distortion removal instrument, our KPI measurement system for ADAS cameras, image sensor characterization according to EMVA 1288 standards, and our camera timing measurement device useful in all camera development applications.
Booth: #707
Image Quality Labs will showcase turnkey embedded camera design and integration solutions, including a camera timing system for validating exposure time and frame rate. Demos feature reference designs, off-the-shelf cameras and ISP tuning, enabling robust AI through high-quality image data.
Booth: #819
Visit our booth to see a presentation of our Benchtop Test Stand for testing cameras at close range. We will demonstrate our image quality testing software, a new test target for measuring information-based dynamic range and other test targets for extracting key performance metrics that correlate with machine vision performance.
Booth: #907
Camera Expert is a simulation-driven digital twin for end-to-end vision systems. From application requirements through scenes, optics, sensors, ISP and AI, the full pipeline is modeled to compare designs and predict image quality and AI performance, enabling reduced risk, better alignment and faster decisions before prototypes.
Booth: #804
Discover our industry-leading low-power, small form factor FPGAs and solutions for automotive, industrial and security applications at the edge. Come see our technologies in action—from cloud to sensor, including low-power FPGAs, solution stacks, software tools, partner solutions and more.
Booth: #416
Lincode’s LIVIS transforms quality control with AI-powered visual inspection. Its no-code platform enables rapid model training, workflow orchestration and analytics, while Edge+ delivers real-time, on-site inspection. Scalable and adaptive, LIVIS provides fast, precise defect detection across automotive, aerospace and electronics manufacturing.
Booth: #815
Living Optics makes hyperspectral imaging practical and accessible for machine vision. Our integrated cameras and software capture spectral data beyond human sight and RGB imaging, enabling AI-driven insights across food, healthcare, security, academia and industry. An Oxford spinout, the company is headquartered in the UK, with growing global reach.
Macnica Americas delivers integrated solutions from sensors to systems, enabling edge AI and embedded vision innovation. Through our Capture, Process, Communicate platform, we help accelerate the development of intelligent products with advanced compute, connectivity and imaging technologies, reducing complexity and speeding time to market.
MagikEye is a 3D sensing company developing ILT (Invertible Light Technology), a depth perception approach that shifts computation from digital processing into the optical domain. By combining encoded illumination with standard CMOS image sensors, ILT enables real-time depth extraction with reduced compute, power, system cost and memory requirements. Unlike stereo, structured light or time-of-flight, ILT minimizes system complexity, enabling scalable deployment of efficient 3D vision across robotics, automotive, consumer and physical AI platforms.
Booth: #316
Smarter embedded vision begins with the right foundation. We’ll be showcasing how our power-efficient FPGAs and secure interfaces enable real-time vision across edge AI, automotive and industrial applications. Visit our booth to experience face recognition with liveness detection, high-speed SLVS-EC and CoaXPress vision bridges and low-latency sensor pipelines for edge AI.
Booth: #302
Micron will showcase how advanced memory and storage solutions enable the next generation of edge AI, computer vision and intelligent robotics. At our booth, we’ll highlight the critical role of high‑performance, power‑efficient memory in meeting the growing demands of vision‑based AI workloads running at the edge.
Booth: #303
MIPS offers a software-first approach to embedded AI. The MIPS Atlas S8200 RISC-V NPU enables physical AI workloads for autonomous platforms. Atlas Explorer is a simulation-based software/hardware co-design platform for Atlas IP core models and physical AI platforms, delivering earlier insights into workloads.
Booth: #805
ModelCat is an autonomous model builder that uses AI to deliver production-ready vision models for a wide range of edge silicon devices. Hand it a labeled dataset or working model along with your requirements for power, speed, memory and accuracy, and ModelCat does the rest. Stop by our booth to see the future of model development.
Booth: #900
Experience our cutting-edge lineup featuring solid-state LiDAR for immersive wide-angle imaging and high-performance drone AF modules for precision aerial focus. We will also present our integrated ToF + RGB sets, delivering superior depth sensing and color accuracy, alongside our secure face recognition solution. These innovations exemplify our commitment to providing high-quality, AI-integrated camera modules that empower smart connectivity and security across diverse industries.
Booth: #509
Nota AI is a leading innovator in AI optimization and compression. Its proprietary AI model optimization platform, NetsPresso, enables high-performance AI models to run efficiently on resource-constrained edge devices through automated optimization and deployment. Nota Vision Agent (NVA), a VLM-powered video monitoring solution, delivers real-time insights across industries, while VLA advancements extend AI from perception to real-world action toward physical AI.
Booth: #408
OpenCV is the world’s biggest computer vision library, downloaded over 1 million times per day. Come see an OpenCV 5 preview, grab a handful of free stickers and enjoy some open-source friendship.
Booth: #909
OpenMV will showcase the OpenMV Cam AE3 and N6—compact, low-power AI vision cameras that run Python on high-performance microcontrollers. See real-time object detection, tracking and edge AI demos running fully on-device, enabling smart, battery-powered robots, industrial sensors and embedded vision systems.
Booth: #809
Optopax designs and manufactures custom camera modules, application-specific optics and multimodal imaging systems. We specialize in sensor-optics co-design, integrating visible, NIR and thermal imaging into compact vision platforms for robotics, industrial inspection, medical devices, subsea and AI-driven edge compute systems.
Booth: #908
We built Avocado OS to make scaling computer vision easy—production-ready Linux with over-the-air (OTA) updates, secure boot and fleet management built in. Pre-validated CUDA, TensorRT and OpenCV across NVIDIA Jetson, NXP and Qualcomm mean your team skips BSP integration entirely. Avocado OS decouples the OS layer from your application logic so ML engineers iterate on models and inference pipelines, not kernel drivers and meta-layers. Go from EVK to production deployment in weeks, not months.
Booth: #609
Quadric simplifies edge AI—one programmable core, lower power, faster innovation. Quadric licenses an innovative semiconductor processor IP core for efficient on-device AI inference. The Chimera general-purpose NPU (GPNPU) is a Python/C++ programmable processor with the flexibility of a DSP and the efficiency of a systolic array matrix engine. Chimera processors span a performance range from 1 TOPS to over 800 TOPS, with design wins in a wide gamut of markets—from automotive to edge LLM server to office automation.
Booth: #821
Qualitas Semiconductor delivers high-performance interface IP solutions, including SerDes, PHY and protocol controllers for PCIe, Ethernet, USB, MIPI and Display standards. Our IP emphasizes low-power, robust signal integrity, comprehensive verification and seamless SoC integration, enabling scalable connectivity across data center, AI, mobile and automotive applications.
Booth: #602
Rapidflare will showcase product intelligence-powered AI agents for the electronics industry that accelerate product selection, customer responses and technical sales by turning complex technical documentation into accurate, source-backed answers and workflows.
Booth: #901
Renesas is pleased to present our RZ/V scalable vision AI microprocessors with AI performance up to 80 TOPS and our RA8 microcontroller with the latest Arm Cortex-M85 core with Helium technology. Visit our booth to see our latest RZ/V Preferred Partner SoMs and SBCs in addition to our Renesas Ready Partner AI applications.
Booth: #621
Squint Insights Studio is a software platform bringing explainability and oversight to AI applications. With our studio, developers iterate more quickly, identify where to improve their models and datasets, cut project costs by up to 40% and deploy pipelines automating error detection and correction, increasing overall trust and safety.
Booth: #617
Stereolabs, a global leader in embedded vision, combines ZED cameras, AI compute and software into a full perception stack, delivering real-time depth sensing, tracking and 3D mapping for robotics and autonomous systems.
Booth: #800
Experience ST’s latest embedded vision innovations: our smart rearview camera with real-time object detection and counting for smarter, safer mobility; our latest time-of-flight (ToF) product, the VL53L9CX dToF 3D mini-LiDAR module, delivering high-resolution depth maps for robotics, AR and more; and our ultra-low-power smart doorbell featuring on-device edge AI for presence and parcel detection—all powered by the latest STM32N6 MCU.
Booth: #502
Sunex is a leading provider of lenses and modules focused on robotic, embedded and computer vision for all market verticals, including extreme environmental solutions, new dual-channel/stereo lens technology, interchangeable lens technology and auto-focus lens options, in addition to new imaging lens offerings for today’s latest sensors.
Booth: #708
Synetic provides world-class data and computer vision solutions that help organizations build, train and deploy AI systems for real-world applications. We support customers with the data, tools and technical expertise needed to improve vision-based AI performance across a range of industries.
Booth: #309
Synopsys ARC NPX6 NPU IP addresses the demands of real-time compute with ultra-low-power consumption for AI applications. To accelerate application software development for the ARC NPX6 NPU IP, the ARC MetaWare MX Development Toolkit provides a comprehensive compilation environment with automatic neural network algorithm partitioning to maximize resource utilization.
Booth: #616
The Ubicept Toolkit unlocks higher quality, more reliable visual data for autonomous systems. Using a physics-based approach, it transforms raw sensor data from SPAD and conventional CMOS cameras into perception-ready outputs that preserve detail in low-light, high-contrast and fast-motion scenes.
Booth: #404
Ultralytics is a leading force in vision AI, best known for our YOLO models. With over 125K GitHub stars, 225M+ Ultralytics downloads and 2B+ model usages, we are the end-to-end platform to annotate, train and deploy production-ready computer vision models.
Booth: #905
AI vision systems often face high compute costs, redundant decoding and heavy bandwidth use when running multiple models. Our demo shows how SMPTE VC-6 solves these issues by enabling multi-model inference from a single stream with shared decoding and region-of-interest refinement, delivering lower compute, reduced bandwidth and scalable real-time performance for edge applications.
Booth: #709
VeriSilicon offers six in-house processing IPs—GPU, NPU, VPU, DSP, ISP and display—plus 1,700+ analog, mixed-signal, RF and interface IPs. Through our SiPaaS model, we serve diverse markets, including consumer electronics, automotive, computing, industrial, data processing and IoT, supporting customers such as fabless firms, IDMs, OEM/ODMs, internet companies and cloud providers.
Booth: #508
Visage Technologies provides specialized computer vision consulting, porting lightweight models to embedded hardware like NVIDIA Jetson and NXP via custom engines. Our Jetson Orin AGX demo showcases crowd analysis, coupling NVIDIA people counting with proprietary face analytics built on visage|SDK to measure real-time engagement and demographics.
Booth: #802
Visidon showcases AI-powered RAW denoise for superior low-light quality, video face super-resolution for sharper facial details in video streams and advanced image enhancement for improved traffic analytics accuracy in challenging lighting and weather conditions.
Booth: #604
We’ll be showcasing our industrial-grade VC MIPI camera modules with MIPI CSI-2 interface, including the new VC MIPI IMX454 for multispectral imaging and the VC MIPI Multiview Cam with nine sensors. Also on display: the plug-and-play VC MIPI Bricks system and embedded vision solutions with onboard AI processing.
Booth: #620
Experience the world’s first fully software-based, end-to-end ISP pipeline powered entirely by AI. Running in real time, Visionary.ai’s ISP enhances video in extreme low light and HDR, boosting sharpness, color fidelity and downstream detection accuracy through a complete neural replacement of traditional ISP blocks.
Booth: #706
Winbond’s extensive product portfolio includes Code Storage Flash and TrustME Secure Flash, alongside Customized Memory Solutions (CMS). These products are widely trusted by Tier 1 customers across industries such as communications, consumer electronics, automotive, industrial applications and computer peripherals.
Booth: #700
YAZAKI Innovations provides high-quality training data and advanced image annotation for edge AI and embedded vision systems. We support automotive and autonomous applications with expertise in 2D/3D data, LiDAR, low-light and infrared environments.
Booth: #906
AMD is redefining intelligent edge and physical AI by bringing data center-class performance to embedded platforms without compromise. We will demonstrate how our ROCm open-source stack unlocks the full compute potential of Radeon iGPUs for real-time AI/ML on power-constrained hardware. Live demos will showcase vision and LLM workloads running locally on Ryzen AI, highlighting an open, flexible stack for deploying any model topology from vision to LLMs at the edge.
Booth: #516
Boston.AI delivers advanced machine vision and edge AI for embedded systems. Experience machine vision combining LiDAR and camera data for precise spatial awareness, sensor fusion and interactive insights on-device without cloud dependency. Explore our AI photo kiosk, transforming live headshots into AI-generated art with real-time image processing and a touchscreen interface.
Booth: #417
Cadence Tensilica is a global leader in configurable digital signal processor (DSP) IP. Based on the versatile Xtensa architecture, developers can generate customized processors for AI, audio, computer vision, radar and many more applications. Key offerings include the HiFi DSP for audio, voice and speech; the Vision DSP for imaging and radar; and the AI Platform for on-device AI processing.
Booth: #402
Efinix is an innovator in efficient FPGA technology. With three product families ranging from 4K logic elements (LEs) up to 1M LEs, our FPGAs span the entire market from entry level to high-end applications. The flagship Titanium family features products with hardened quad-core RISC-V processors, LPDDR4x controllers and high-speed SerDes. The free downloadable tools provide an intuitive and high-performance path from HDL to bitstream with integrated C/C++ flows for the RISC-V and turnkey AI acceleration.
Booth: #503
eInfochips, an Arrow company, provides solutions across the product stack—from edge to cloud. We will be showcasing our niche machine vision and edge AI solutions, developed leveraging 30+ years of engineering excellence, reference designs and framework based on the latest platforms.
Booth: #902
Expedera provides customizable neural engine semiconductor IP that dramatically improves performance, power and latency while reducing cost and complexity. Expedera’s NPU solutions are scalable and produce superior results in edge nodes and smartphones to automotive applications.
Booth: #520
Originating from a lineage of advanced image and signal processing, FotoNation’s sixth-generation AI IP empowers low-power, low-latency and high-performance sensor compute. These IPs are meticulously crafted for the contemporary era of “companion and physical AI,” characterized by perpetual human-like perception. Stop by our booth for the unveiling.
Booth: #807
Discover our industry-leading low-power, small form factor FPGAs and solutions for automotive, industrial and security applications at the edge. Come see our technologies in action—from cloud to sensor, including low-power FPGAs, solution stacks, software tools, partner solutions and more.
Booth: #416
MagikEye is a 3D sensing company developing ILT (Invertible Light Technology), a depth perception approach that shifts computation from digital processing into the optical domain. By combining encoded illumination with standard CMOS image sensors, ILT enables real-time depth extraction with reduced compute, power, system cost and memory requirements. Unlike stereo, structured light or time-of-flight, ILT minimizes system complexity, enabling scalable deployment of efficient 3D vision across robotics, automotive, consumer and physical AI platforms.
Booth: #316
Smarter embedded vision begins with the right foundation. We’ll be showcasing how our power-efficient FPGAs and secure interfaces enable real-time vision across edge AI, automotive and industrial applications. Visit our booth to experience face recognition with liveness detection, high-speed SLVS-EC and CoaXPress vision bridges and low-latency sensor pipelines for edge AI.
Booth: #302
Micron will showcase how advanced memory and storage solutions enable the next generation of edge AI, computer vision and intelligent robotics. At our booth, we’ll highlight the critical role of high‑performance, power‑efficient memory in meeting the growing demands of vision‑based AI workloads running at the edge.
Booth: #303
MIPS offers a software-first approach to embedded AI. The MIPS Atlas S8200 RISC-V NPU enables physical AI workloads for autonomous platforms. Atlas Explorer is a simulation-based software/hardware co-design platform for Atlas IP core models and physical AI platforms, delivering earlier insights into workloads.
Booth: #805
Nota AI is a leading innovator in AI optimization and compression. Its proprietary AI model optimization platform, NetsPresso, enables high-performance AI models to run efficiently on resource-constrained edge devices through automated optimization and deployment. Nota Vision Agent (NVA), a VLM-powered video monitoring solution, delivers real-time insights across industries, while VLA advancements extend AI from perception to real-world action toward physical AI.
Booth: #408
OpenCV is the world’s biggest computer vision library, downloaded over 1 million times per day. Come see an OpenCV 5 preview, grab a handful of free stickers and enjoy some open-source friendship.
Booth: #909
OpenMV will showcase the OpenMV Cam AE3 and N6—compact, low-power AI vision cameras that run Python on high-performance microcontrollers. See real-time object detection, tracking and edge AI demos running fully on-device, enabling smart, battery-powered robots, industrial sensors and embedded vision systems.
Booth: #809
We built Avocado OS to make scaling computer vision easy—production-ready Linux with over-the-air (OTA) updates, secure boot and fleet management built in. Pre-validated CUDA, TensorRT and OpenCV across NVIDIA Jetson, NXP and Qualcomm mean your team skips BSP integration entirely. Avocado OS decouples the OS layer from your application logic so ML engineers iterate on models and inference pipelines, not kernel drivers and meta-layers. Go from EVK to production deployment in weeks, not months.
Booth: #609
Quadric simplifies edge AI—one programmable core, lower power, faster innovation. Quadric licenses an innovative semiconductor processor IP core for efficient on-device AI inference. The Chimera general-purpose NPU (GPNPU) is a Python/C++ programmable processor with the flexibility of a DSP and the efficiency of a systolic array matrix engine. Chimera processors span a performance range from 1 TOPS to over 800 TOPS, with design wins in a wide gamut of markets—from automotive to edge LLM server to office automation.
Booth: #821
Synetic provides world-class data and computer vision solutions that help organizations build, train and deploy AI systems for real-world applications. We support customers with the data, tools and technical expertise needed to improve vision-based AI performance across a range of industries.
Booth: #309
Synopsys ARC NPX6 NPU IP addresses the demands of real-time compute with ultra-low-power consumption for AI applications. To accelerate application software development for the ARC NPX6 NPU IP, the ARC MetaWare MX Development Toolkit provides a comprehensive compilation environment with automatic neural network algorithm partitioning to maximize resource utilization.
Booth: #616
VeriSilicon offers six in-house processing IPs—GPU, NPU, VPU, DSP, ISP and display—plus 1,700+ analog, mixed-signal, RF and interface IPs. Through our SiPaaS model, we serve diverse markets, including consumer electronics, automotive, computing, industrial, data processing and IoT, supporting customers such as fabless firms, IDMs, OEM/ODMs, internet companies and cloud providers.
Booth: #508
Learn about the newest speakers, sessions and other noteworthy details about the Summit Program by leaving us a few details.

Interested in sponsoring or exhibiting?
The Embedded Vision Summit gives you unique access to the best qualified technology buyers you’ll ever meet.
Want to contact us?
Use the small blue chat widget in the lower right-hand corner of your screen, or the form linked below.
STAY CONNECTED
Follow us on X and LinkedIn.