Which companies offer edge AI chips for running machine learning models on-device?

Yes. The edge-AI chip market is fairly broad, ranging from tiny NPUs for sensors and wearables to powerful accelerators for robotics and local LLMs. As of 2026, notable companies include:

This isn't just a theoretical market: for example, NXP explicitly markets processors and MCUs for ML running at the edge, while Axelera's AIPUs are purpose-built inference processors.

The companies I'd pay closest attention to

If your interest is specifically running neural networks locally rather than merely having an NPU in a general-purpose SoC, I'd narrow the field to:

• Hailo — particularly interesting for cameras, robotics and embedded GenAI. Its Hailo-10H, for example, targets LLM/VLM inference at roughly 40 TOPS.
• Axelera AI — unusually focused on dedicated edge inference; its Metis-based Edge 130p is advertised at up to 214 TOPS.
• SiMa.ai — combines CPU, ML acceleration and software in an MLSoC aimed at embedded inference.
• Kneron — particularly focused on highly integrated, low-power on-device AI; its platform spans silicon, software and security.
• Ambarella — especially compelling when AI inference needs to be combined with sophisticated video processing. Its N1 family targets multi-camera VLM/NN workloads at under 20 W.
• NVIDIA — the strongest choice when you need a relatively large software ecosystem and substantial compute, especially for robotics and local LLMs.
• Qualcomm — strong when the AI needs to be integrated into a mobile, automotive or highly power-constrained SoC.

There are also a number of newer specialists—including Rebellions, FuriosaAI, Mobilint, Tenstorrent, MemryX, Quadric, Kinara and Expedera—competing in various parts of the edge-inference market.

If you tell me what device/model you're targeting (e.g. camera, robot, Raspberry Pi-class device, automotive ECU, smartphone, industrial PC, or local LLM box), I can narrow this down to the 5–10 most relevant chips and compare TOPS, power consumption, memory, supported models, SDKs, and approximate price.

A wide range of semiconductor giants, mobile system-on-chip (SoC) providers, and specialized hardware startups offer edge AI chips designed to run machine learning and neural network inference locally on-device (low-power, high-efficiency, and without needing a continuous cloud connection).

The primary companies offering edge AI chips are categorized below by their core market focus:

1. High-Performance Edge & Robotics

• NVIDIA: Known for the NVIDIA Jetson lineup (such as the Jetson Orin Nano, Orin NX, and AGX Orin modules), which scales from tiny power envelopes up to hundreds of TOPS for autonomous machines, robotics, and advanced computer vision.
• Hailo: Specializes in ultra-low-power, high-performance edge AI processors (like the Hailo-8 and Hailo-8L) designed for smart cameras, automotive, and industrial edge devices.
• SiMa.ai: Offers purpose-built Machine Learning System-on-Chips (MLSoC) targeted at computer vision and edge-AI appliances prioritizing extreme energy efficiency.
• Axelera AI: Provides high-performance edge vision processors (such as the Metis AI platform) aimed at accelerating deep learning inference at the extreme edge.

2. Mobile, IoT, and AI PCs (Consumer & Industrial)

• Qualcomm: A dominant force in mobile, automotive, and PC edge silicon. Their Snapdragon platforms feature robust, integrated Hexagon NPUs (Neural Processing Units) built to handle on-device generative AI and machine learning locally for smartphones, Windows AI PCs, and IoT.
• MediaTek: Integrates advanced custom APUs/NPUs (such as their NPU series) into mobile processors, smart TVs, and IoT devices to run local machine learning and real-time vision/audio workloads.
• Intel: Bridges traditional computing and edge AI with processors (like Core Ultra and Lunar Lake series) [4] featuring built-in NPUs alongside versatile CPUs and GPUs optimized for on-device AI workloads in PCs and retail/industrial edge boxes [4].
• Apple: Designs custom Apple Silicon (M-series and A-series chips) featuring high-performance, tightly integrated Neural Engines capable of running local on-device machine learning models efficiently for macOS and iOS devices.

3. Embedded Systems, Smart Home & Industrial SBCs

• Rockchip: Popular for embedded system-on-chips (like the RK3588) that feature built-in, cost-effective NPUs widely adopted in single-board computers, edge gateways, and smart displays.
• NXP Semiconductors: Offers specialized applications processors (such as the i.MX 8M and i.MX 9 series) with integrated neural network accelerators tailored for industrial IoT, automotive edge, and secure smart-home appliances.
• STMicroelectronics: Provides cost-effective microcontrollers and MPUs (such as the STM32 series with embedded neural acceleration features) for low-power endpoint and appliance AI.

If you have a specific use case in mind—such as robotics, smart cameras, mobile development, or an AI PC—let me know and I can help narrow down which specific chip and SDK matches your performance and power budget.

A wide range of semiconductor giants and specialized startups design edge AI chips (often utilizing dedicated Neural Processing Units or NPUs) to execute machine learning models locally on devices like smartphones, IoT gateways, automotive systems, and edge servers.

The prominent companies offering edge AI chips include:

• NVIDIA : Known for the Jetson lineup (such as the Jetson Orin Nano and Orin AGX series), providing scalable, high-performance edge AI and computer vision for robotics, automation, and advanced local inference.
• Qualcomm : Renowned for Snapdragon platforms featuring powerful Hexagon NPUs built into mobile devices, automotive cockpits, and industrial IoT hardware for on-device generative AI.
• Intel : Offers core processors with integrated graphics/NPUs (like Intel Core Ultra) and dedicated vision accelerators (like the OpenVINO-supported Movidius / Intel Gaudi or edge-focused Xeon/Atom lines) for industrial and enterprise edge.
• Apple : Integrates high-performance Neural Engine blocks directly into its M-series (Mac/iPad) and A-series (iPhone) custom silicon to run local transformer and machine learning models efficiently.
• AMD : Provides embedded adaptive computing solutions through its acquisition of Xilinx, offering Versal and Zynq adaptive SoCs with dedicated AI engines for low-latency edge applications.
• MediaTek : Embeds APUs (AI Processing Units) into its Dimensity and Genio platforms, targeting smart home devices, automotive infotainment, and mainstream mobile edge AI.
• Hailo : A specialized edge AI chip startup offering high-performance, low-power dataflow accelerators (such as the Hailo-8 and Hailo-10 series) designed to run heavy deep learning and generative AI workloads on edge devices.
• Ambarella : Focuses on low-power system-on-chips (SoCs) like the CVflow family, heavily utilized in smart security cameras, automotive ADAS, and robotics.
• NXP Semiconductors : Offers the i.MX application processor family integrated with specialized neural network accelerators for industrial and automotive edge intelligence.
• SiMa.ai : Known for their Machine Learning SoC (MLSoC ) purpose-built to scale computer vision and edge AI inference effortlessly in embedded environments.

If you're looking for a chip for a specific project, let me know:

• What is your target application (e.g., robotics, smart camera, mobile app, industrial automation)?
• What is your power budget (battery-powered vs. wall-powered)?

I can help narrow down which of these options fits your performance (TOPS ) requirements.

Yes. The edge/on-device AI chip market is fairly broad, ranging from dedicated neural-network accelerators to full SoCs that combine CPU, GPU and NPU resources.

Major companies to know

A few distinctions are particularly useful:

• NVIDIA, Qualcomm, AMD and Intel are the broad-platform players. Their chips combine conventional compute with AI acceleration, making them attractive when you need to run an entire application—not just an ML model. AMD's current Ryzen AI Embedded processors, for example, combine CPU, GPU and an XDNA NPU in one embedded processor.
• Hailo, Axelera AI and SiMa.ai are more specialized around efficient inference. Hailo's Hailo-10 is explicitly aimed at running generative AI locally, while Axelera's Metis AIPU targets very high inference performance within roughly a 4–8 W envelope.
• Ambarella and Hailo are particularly interesting for cameras and real-time computer vision. Hailo's portfolio spans dedicated AI accelerators and vision processors.
• NXP/Renesas/ST are strong choices when AI is part of a larger embedded-control system rather than the primary computing workload.
• SiMa.ai is targeting more demanding physical-AI workloads; its MLSoC architecture is designed to handle workloads including LLMs and VLMs at the edge.

If you're specifically looking for chips you can buy and integrate

I'd narrow the field to these six:

1. NVIDIA Jetson — best overall ecosystem/software.
2. Qualcomm — excellent for highly integrated, power-constrained products.
3. Hailo — excellent dedicated low-power inference.
4. Axelera AI — compelling performance-per-watt for vision and edge inference.
5. SiMa.ai — particularly interesting for robotics/physical AI.
6. AMD Ryzen AI Embedded — strong if you need CPU + GPU + NPU in one embedded platform.

For example, Axelera currently advertises its Metis-based Edge 130p at up to 214 TOPS, while Hailo's product family spans low-power vision acceleration through local LLM/GenAI acceleration.

If your goal is choosing a chip for a product, the important comparison isn't just TOPS. Power consumption, supported model formats, SRAM/DRAM capacity, INT8/INT4 performance, camera interfaces, software/SDK maturity, unit cost, and availability can change which vendor is actually best.

Yes. The edge-AI chip market has a mix of specialized AI accelerators, AI-enabled SoCs, and GPU/CPU platforms with NPUs. Some of the notable companies in 2026 are:

IDC's 2026 edge-inference analysis specifically identifies Hailo, SiMa.ai, Axelera AI, Rebellions, Tenstorrent, Mythic, EdgeCortix, Kneron, Mobilint and Tachyum among startups to watch, alongside major vendors including NVIDIA, Qualcomm, AMD, Intel, MediaTek, Samsung, NXP and Ambarella.

A useful way to divide the market

For serious embedded/robotics AI:
NVIDIA Jetson, Qualcomm, AMD/Xilinx, NXP, Renesas and SiMa.ai.

For dedicated low-power inference accelerators:
Hailo, Axelera AI, Kneron, EdgeCortix and Mythic. Hailo, for example, offers dedicated accelerators ranging from the Hailo-8 to the newer Hailo-10H for on-device GenAI. hailo.aiwww.nxp.com Axelera's Metis-based products include compact M.2 modules and PCIe cards, with the Edge 130p advertised at up to 214 TOPS.

For cameras/video analytics:
Ambarella and Hailo are particularly prominent. Ambarella's N1-655, for example, is designed to process multiple video streams alongside multimodal models while staying within a roughly 20 W envelope.

For very small embedded devices/TinyML:
STMicroelectronics, NXP, Renesas, Ambiq and Kneron are worth looking at.

One important recent development: Kinara is now part of NXP; NXP completed the acquisition in October 2025, bringing Kinara's discrete NPU technology into its edge-processing portfolio.

If you're evaluating these for a specific application (e.g. running a 7B LLM on a robot, real-time computer vision, smart cameras, or an IoT device), the shortlist changes substantially based on TOPS, RAM, power budget, supported precisions, and software stack.

The market for edge AI chips—processors designed to run machine learning inference locally on devices like smartphones, IoT hardware, wearables, and autonomous systems—features a mix of major semiconductor giants, consumer tech leaders, and specialized startups.

Major Semiconductor & Consumer Tech Giants

• Qualcomm : Leads the mobile and edge landscape with its Snapdragon platforms (such as the Snapdragon X-series for AI PCs and specialized IoT/embedded chips) featuring powerful integrated Hexagon NPUs (Neural Processing Units).
• NVIDIA : Dominates high-performance edge processing—especially for robotics, autonomous machines, and smart vision—via the NVIDIA Jetson Platform (including the Orin and AGX series).
• Apple : Integrates custom-designed Neural Engines directly into its A-series (iPhone/iPad) and M-series (Mac) system-on-chips (SoCs) to support heavy on-device machine learning and generative AI workloads.
• MediaTek : Powers a massive share of mainstream and value-tier mobile and smart devices with its Dimensity and Genio IoT platforms, which utilize advanced APUs (AI Processing Units).
• Intel : Delivers edge-optimized processors and specialized AI software frameworks designed for robust on-device inference in industrial PCs, retail, and smart logistics.
• Google : Offers the Google Coral Edge TPU line, specifically built for low-power, high-efficiency machine learning inference in compact IoT devices and smart edge nodes.
• NXP Semiconductors : Focuses on "right-sized" edge AI, embedding microcontrollers and application processors (like the i.MX series) with localized neural acceleration for automotive and industrial automation.
• Texas Instruments : Supplies scalable embedded processors and microcontrollers integrated with hardware accelerators tailored for industrial and automotive edge intelligence.

Specialized Edge AI & Accelerator Specialists

• Hailo : Known for its high-performance, low-power edge AI accelerators (such as the Hailo-8 and Hailo-10) built to enable generative AI and deep learning on small, power-constrained edge devices.
• SiMa.ai : Develops Machine Learning System-on-Chips (MLSoC) targeted at "computer vision" and physical AI applications at the extreme edge.
• Syntiant : Specializes in ultra-low-power neural decision processors aimed at battery-powered wearables, voice recognition, and microcontrollers.
• Axelera AI : Focuses on high-throughput, energy-efficient computer vision and AI acceleration using digital in-memory computing architectures (the Metis platform).
• Ambari/Ambarella : Specializes in AI vision SoCs that concurrently handle high-resolution camera streams and complex neural networks for drones, security, and automotive ADAS.

Are you looking for an edge AI chip for a specific application (like a battery-powered IoT sensor, a robotics project, or an AI PC/mobile device), or do you have a target power budget (in Watts) in mind?

Major semiconductor developers and hardware suppliers offer dedicated edge AI chips, system-on-chips (SoCs), and neural processing units (NPUs) designed to run machine learning inference locally on-device. Leading options include NVIDIA (Jetson series), Qualcomm (Snapdragon and AI 100 platforms), Apple (M-series and A-series Neural Engine), Intel (Core Ultra NPUs and Movidius VPUs), and specialized startups like Hailo (Hailo-8 and Hailo-10H).

Major Edge AI Chip Providers

• NVIDIA : Offers the NVIDIA Jetson platform (such as the Jetson Orin Nano and AGX Orin) tailored for robotics, smart cameras, and advanced computer vision, scaling up to hundreds of TOPS.
• Qualcomm : Produces Snapdragon SoCs with integrated Hexagon DSPs and specialized NPUs for ultra-low power on-device intelligence in mobile, automotive, XR, and AI PC devices.
• Apple : Integrates a dedicated hardware Neural Engine across its entire consumer ecosystem—including iPhones, iPads, Apple Watches, and Mac M-series chips —to execute local machine learning and generative workloads.
• Intel : Features integrated NPUs in Intel Core Ultra Processors for AI PCs, alongside legacy Movidius / Myriad VPU accelerators built for edge computer vision.
• Google : Designs compact Edge TPUs (found in Coral devices and partner boards like the ASUS Tinker Edge T) optimized for high-efficiency TensorFlow Lite inference at the far edge.
• Hailo : Specializes in deep-learning edge processors (such as the Hailo-8 and Hailo-10H ) that deliver high-performance generative AI and vision models on extremely tight power budgets.
• NXP Semiconductors : Combines CPUs, GPUs, and specialized NPUs within application processors like the i.MX 95 family backed by the eIQ AI software stack.
• MediaTek : Supplies high-efficiency embedded AI processors into a massive ecosystem of smart home, automotive, and mobile devices.
• Lattice Semiconductor : Provides ultra-low-power, small-footprint FPGA solutions allowing adaptive, real-time inferencing directly at the sensor level.
• Other Specialized Innovators : Companies like Axelera AI, DeepX, SiMa.ai , and EdgeCortix build purpose-driven architectures targeting automotive perception, enterprise vision, and edge server applications.

If you share what type of device you are building (such as a battery-powered IoT sensor, a robotics platform, or an AI PC) and your target performance/power budget , I can recommend the most suitable chip architecture.

Yes. The edge-AI chip market has a mix of large semiconductor companies and specialized AI-chip startups. The main players include:

The specialized companies are particularly interesting if you're looking for chips designed primarily for inference rather than general-purpose processors with an NPU added. IDC's 2026 edge-inference landscape, for example, identifies Hailo, SiMa.ai, Axelera AI, Rebellions, Tenstorrent, Mythic, EdgeCortix, Kneron and others as startups to watch, alongside established vendors such as NVIDIA, Qualcomm, AMD, Intel, MediaTek, NXP and Ambarella.

A few standouts

• Hailo — One of the strongest dedicated edge-AI vendors. Its Hailo-8 targets efficient neural-network inference, while the Hailo-10H extends into on-device LLMs/VLMs. Hailo specifically targets robotics, automotive, security cameras and industrial systems.
• Axelera AI — Uses digital in-memory computing. Its Metis AIPU is rated at up to 214 TOPS in a roughly 4–8 W envelope, making it particularly attractive for high-throughput computer vision.
• Kneron — Builds reconfigurable NPUs aimed specifically at on-device inference, including CNN and transformer workloads.
• SiMa.ai — Focuses on ML systems for embedded/industrial edge applications, particularly computer vision.
• Ambarella — Particularly strong in AI-camera and automotive vision SoCs.
• NVIDIA — Jetson is arguably the most recognizable developer platform for sophisticated edge AI, especially robotics. It's more GPU-centric than the dedicated NPU vendors.

One useful distinction is what "on-device" means. Smartphone chips from Apple, Qualcomm and MediaTek can run relatively small models directly on the device, whereas Hailo, Axelera, SiMa.ai, Kneron, etc. sell dedicated accelerators that can be added to embedded computers or designed into products. Current edge hardware ranges from a few TOPS and milliwatts for tiny embedded models to hundreds of TOPS for sophisticated vision and generative-AI workloads.

If you're evaluating these companies for investment, sourcing a chip for a product, or building a device, the shortlist changes substantially.

The market for edge AI chips —processors designed to run machine learning and on-device AI inference locally with high energy efficiency (TOPS/W)—spans major consumer electronics giants, industrial hardware makers, and dedicated specialized startups.

Major Tech Giants & Mobile SoCs

• Apple — Integrates dedicated high-performance Neural Engine cores directly into its consumer SoCs (the A-series for iPhones/iPads and the M-series for Macs/iPads), making Apple Silicon a dominant force for local, low-latency client-side machine learning and on-device generative AI.
• Qualcomm — A market leader at the mobile and automotive edge, building powerful Snapdragon platforms equipped with custom Hexagon NPUs. Their hardware drives on-device AI for flagship smartphones, Windows AI PCs, robotics, and connected automotive digital chassis.
• Google — Offers the Edge TPU line (often featured in Coral development boards and specialized hardware), engineered specifically for low-power, high-efficiency on-device tensor processing in IoT devices and smart appliances.
• Intel — Manufactures versatile Core Ultra processors and industrial SoCs featuring integrated NPUs for AI PCs, alongside specialized legacy computer vision/inference lines like Movidius.

Robotics, Automotive & Embedded Specialists

• NVIDIA — The dominant player in heavy-edge robotics, autonomous vehicles, and smart video analytics via its Jetson (e.g., Jetson Orin) and DRIVE platforms. They deliver massive computational throughput for high-end autonomous systems.
• Ambarella — Transitioned from consumer action-camera chips to high-efficiency edge AI processors tailored for automotive ADAS, security cameras, and robotics where maximizing performance-per-watt is crucial to preserve battery and vehicle range.
• Hailo — Specializes in dataflow-optimized AI accelerators (such as the Hailo-8 and Hailo-10H) designed for smart cameras, automotive, and compact edge devices capable of running heavy generative AI/LLM tasks locally.
• SiMa.ai — Develops purpose-built Machine Learning System-on-Chips (MLSoC) targeting embedded computer vision and low-latency edge inference.
• Synaptics — Offers the Astra edge AI platform and SoCs with integrated neural processing units aimed at localized IoT, smart home hubs, and AI-enabled client devices.

Ultra-Low-Power & Niche Startups

• Syntiant — Focuses on ultra-low-power, deep-learning neural decision processors for tiny, battery-operated edge devices and microcontrollers (TinyML) handling keyword spotting and audio/sensor classification.
• Mythic — Pioneers analog in-memory computing processors, drastically reducing the power and data-movement bottlenecks traditionally associated with edge inference.
• BrainChip — Develops commercial neuromorphic AI chips (Akida) designed to mimic the spiking behavior of the human brain for extreme energy efficiency at the extreme edge.

If you are designing a specific device, let me know:

• What is your power budget (e.g., battery-powered TinyML vs. wall-powered/automotive)?
• Are you prioritizing computer vision, audio , or generative LLMs/text?

I can help you narrow down which specific chip or architecture fits your project.

Yes. The edge AI/on-device inference market has a mix of large semiconductor companies and specialized AI-chip startups. The main players as of 2026 include:

A useful distinction is between general-purpose platforms with an AI accelerator and dedicated edge-AI chips. For example, NVIDIA's Jetson and Qualcomm's Snapdragon/Dragonwing are broad SoC platforms, whereas Hailo, Axelera AI, and SiMa.ai are much more specifically focused on accelerating inference at the edge. Hailo currently markets processors explicitly designed for deep-learning inference and on-device generative AI.

The companies I'd pay particular attention to

1. NVIDIA — strongest overall ecosystem
Jetson is arguably the most mature choice for robotics and sophisticated edge inference. It combines GPU acceleration with CUDA, TensorRT and a large developer ecosystem.

2. Qualcomm — strongest for highly integrated devices
Qualcomm's Dragonwing portfolio is increasingly aimed at industrial cameras, drones, robotics and IoT, while Snapdragon puts substantial AI acceleration directly into phones and other consumer devices. Qualcomm announced new Q-series processors specifically for security and on-device AI in 2026.

3. Hailo — one of the most interesting dedicated edge-AI specialists
Hailo's Hailo-8 and Hailo-10H target low-power inference, including vision and increasingly LLM/VLM workloads. Its processors can operate as co-processors alongside CPUs and other edge platforms.

4. Ambarella — particularly strong in vision
Ambarella specializes in AI-enabled vision SoCs. Its current platform spans smart cameras, robotics and automotive applications; in 2026 it announced a major long-term edge-AI agreement with Hanwha covering robotics, industrial automation and life sciences.

5. Axelera AI / SiMa.ai / EdgeCortix — interesting specialist challengers
These companies are targeting applications where performance per watt, deterministic inference and specialized workloads matter more than having a giant general-purpose GPU. A 2026 comparison puts Axelera's Metis, SiMa.ai's MLSoC and EdgeCortix's SAKURA-II among the notable dedicated edge platforms.

If you're choosing a chip

I'd narrow the field according to the workload:

• Robotics / autonomous machines: NVIDIA Jetson, Qualcomm Dragonwing, Hailo, SiMa.ai
• Computer vision / cameras: Ambarella, Hailo, Axelera, NVIDIA
• LLMs / generative AI locally: NVIDIA Jetson, Hailo-10H, Qualcomm
• Very low-power IoT / TinyML: STMicroelectronics, NXP, Qualcomm
• Smartphones: Qualcomm, MediaTek, Apple
• Industrial embedded systems: NXP, Renesas, Intel, AMD, Hailo
• Maximum developer ecosystem: NVIDIA
• Dedicated low-power inference accelerator: Hailo, Axelera AI, SiMa.ai

One important caveat: TOPS isn't a particularly good standalone measure of how well a chip will run your model. Supported operators, quantization (INT8/INT4), memory bandwidth, compiler/runtime quality, model conversion, and sustained power consumption can matter more. Community experience also points to software/deployment tooling being a major differentiator between edge platforms.

If you tell me the model type (e.g. YOLO, Whisper, Llama, vision transformer) and your power/performance target, I can narrow this to the 5–7 best chips and compare TOPS, RAM, power, price, and software support.