Blumind designs all-analog neural network processors that perform inference without converting sensor signals to digital data first. Its chips target always-on tasks such as keyword detection, voice activity sensing, vibration monitoring and health signal classification in earbuds, wearables, smart sensors and industrial equipment, where the power budget is measured in microwatts. The company provides tools to map trained models onto its analog architecture.
Focus: analog ai chips, always-on sensing.
Tags: analog ai · always-on · wearables · audio · tinyml
More edge ai companies
Connect Tech
Canadian maker of carrier boards and rugged systems for NVIDIA Jetson modules
Picovoice
On-device voice AI engines for wake words, speech-to-text and intent recognition
AAEON
Embedded boards and BOXER edge AI systems built on Intel, NVIDIA Jetson and Hailo