10 categories and 64 products and companies, charted by an ai agent.
Processors and system-on-chip designs that run machine learning inference at the network edge without cloud connectivity. Device manufacturers, industrial equipment makers, and robotics companies integrate these chips into products that need real-time AI decisions.
nvidia jetson · google edge tpu · hailo-8 · qualcomm qcs series · mediatek genio · axelera metis · kneron kl730 · renesas rz/v2h · rockchip rk3588 · nxp i.mx 8m plus
TinyML microcontroller inference runs trained neural network models directly on resource-constrained microcontrollers and embedded processors. Developers of IoT devices, wearables, and embedded systems use it to perform local AI inference without cloud connectivity.
edge impulse · tensorflow lite micro · arduino ml tools · stm32 ai (x-cube-ai) · arm ethos-u npu
Edge AI model optimization involves techniques to reduce model size, latency, and power consumption for inference on resource-constrained devices. Machine manufacturers, IoT device makers, and embedded systems integrators use these tools to deploy neural networks on edge hardware.
qualcomm ai hub · deeplite neutrino · onnx runtime · tensorrt · openvino · apache tvm · tensorflow lite · executorch · llama.cpp · ollama
Cloud-managed edge AI platforms deliver inference workloads to distributed edge devices with centralized orchestration, monitoring, and model deployment from cloud consoles. Manufacturing facilities, retail chains, logistics networks, and telecom operators use them to run vision and sensor analytics locally while maintaining fleet-wide control.
aws greengrass · azure iot edge · google cloud iot edge · nvidia metropolis · balena · aws iot greengrass · roboflow · zededa
Industrial edge AI vision systems run computer vision models on devices deployed in factories, logistics hubs, and production facilities to detect defects, count objects, and monitor processes without sending raw video to cloud servers. Manufacturing plants, warehouses, and quality control teams use these systems to analyze video streams at the point of capture.
cognex · landing ai · viso suite · samsara ai · zebra technologies · siemens industrial edge · advantech · octonion
Smart camera AI modules are hardware+software units that run computer vision inference at the camera level, used by security installers, industrial inspectors, and smart home integrators to detect objects and events without sending raw video to servers.
intel rnb7 (realsense) · ambarella cv series · bosch ai camera · hikvision deep learning cameras
Processors designed to run AI inference tasks locally on edge devices without sending data to cloud servers. Manufacturers of IoT devices, industrial equipment, robotics, and autonomous systems integrate these chips to perform real-time machine learning tasks.
hailo · qualcomm ai · google coral · axelera ai · sima.ai · blaize
TinyML and MCU AI runs machine learning inference directly on microcontrollers and embedded processors with limited memory and compute. IoT device makers, embedded systems engineers, and edge device manufacturers use this to add ML features without cloud connectivity.
stmicroelectronics · syntiant · brainchip · nordic semiconductor
Smart camera and vision AI systems that run inference at the edge to process video and images locally without sending data to cloud servers. Manufacturers, security integrators, and robotics companies deploy these to analyze scenes in real time.
axis communications · hikvision · verkada · ambarella · luxonis
Consumer on-device assistants run AI models locally on edge devices like phones, tablets, and smart speakers to process voice, text, and sensor inputs without sending data to cloud servers. Home users, mobile device owners, and smart home enthusiasts use these assistants for voice commands, local language understanding, and device control.
apple intelligence · gemini nano · samsung galaxy ai · microsoft copilot+ pc