BrainCard, putting neurons at work for the Intelligence of Things

BrainCard is a trainable pattern recognition board for IoT and smart appliances featuring a NeuroMem CM1K chip with 1024 trainable neurons interfaced via SPI communication to Arduino and Raspberry Pi processor boards and ready to learn and recognize patterns.

Sensor data collected through Arduino shields and other plug-in modules can be assembled into feature vectors of up to 256 bytes. Their learning can be triggered by external user inputs, programmable time stamps, but also by the detection of novelties made by the neurons themselves. Once trained, the neurons can be continuously interrogated to report known patterns or events, or to report novelty. Depending on your application, the output of the neurons can control actuators, trigger a selective recording or transmission or else. Applications include identification, surveillance, tracking, adaptive control and more.

We are presently out of stock, but our next batch is due early 2017. Please register to our newsletter to be kept posted as soon as we receive them.


  • CM1K chip (1024 neurons)
  • Xilinx Spartan 6 FPGA
  • Audio MEMS
  • 16 MB SDRAM
  • A/D converter


  • Arduino connectors. Only supports Arduino boards with 3.3v IOs.
  • Raspberry Pi connector
  • Intel® Edison connector (and  dedicated USB connector)
  • SD card slot
  • Connector to stack additional CM1K chips and expand the neural network
  • Connector compatible with the RaspiCam camera module
  • Mini HDMI connector
  • Power supply through the USB port or the Arduino power connectors

The BrainCard basic API includes  functions such as Learn pattern, Recognize pattern, Save neurons, Load neurons. SPI Read/Write functions give you full access to the neural network to change the mode of classification (RBF or KNN), activate multiple contexts, and more. Examples are supplied in Arduino and Python.

FPGA programmers can greatly expand the capabilities and speed performance of the BrainCard by programming direct interfaces to sensors such as an external RaspiCam module or the on-board audio Mems, extracting feature vectors in real-time from the source signals, dispatching them to the neurons under different contexts and making a final decision. Other possible expansions include an HDMI display and connectivity to an Intel Edison to turn the BrainCard into a fully embedded system.