Major Applications Leveraging Eta Compute’s Technology
What is Continuous Voltage & Frequency Scaling?
Cifar10 Image Recognition System consumes <1mJ per inference
Always on command recognition consumes 300uA
The ultra-low-power AI silicon startup Eta Compute has production silicon of its first product. The ECM3532 promises to enable AI and ML computation in IoT applications such as sensor nodes for a power budget of just microwatts. -from EE Times
The chip, called the ECM 3532, consumes as little as 100 microwatts and is designed to perform AI-enabled tasks such as object identification, recognize wake words or sounds, and analyze data from a variety of sensors. -from IEEE Spectrum
Software platform and Processor are ideal for artificial intelligence in edge devices with environmental, sound, motion and image sensors.