Pattern Recognition Chip with 1024 Neurons in Parallel: The CogniMem™ CM1K Chip is the first ASIC version of the CogniMem neural network product line optimized for Cognitive Computing. It features 1024 neurons working in parallel implementing two reknown non-linear classifiers. It can recognize patterns at high speed while coping with ill-defined data, unknown events and changes of contexts and working conditions.

Stackable Module (2048 neurons): The CMEnK Module is a stackable module populated with one or two CM1K chips. An 80-pins hirose connector allows communicating with the chips through their parallel bus or serial I2C bus. Up to four modules can be stacked on top of each other thus expanding the size of the network by increments of 1024 or 2048 neurons, up to 8192.

High Performance Pattern Recognition, 3D Stackable Module : The CogniBlox board features four CM1K Chip or a total of 4096 neurons, a Field Programmable Gate Array (FPGA), memory and a variety of connectors to interface with sensors and other devices.

Module: CM1K-PGA69 is an ideal component for prototyping designs to evaluate the capability and performance of the CM1K.Applications include but are not limited to simple real-time signal and/or video learning and recognition. The output of the neurons can be directly connected to indicators such as seven segment displays, LED, speech synthesizer or else.

Module: CM1K-PGA69 is an ideal component for prototyping designs to evaluate the capability and performance of the CM1K.Applications include but are not limited to simple real-time signal and/or video learning and recognition. The output of the neurons can be directly connected to indicators such as seven segment displays, LED, speech synthesizer or else.

CogniPat Toolkit Pattern Recognition Development Kit

Starter Kit: Evaluation board for video and image recognition applications featuring one CM1K Chip, one Aptina video sensor, a Field Programmable Gate Array (FPGA), a USB connector and variety of I/Os.

Prototyping Module: Prototyping module featuring one CM1K chip and ready to plug into a standard breadboard. Easy to connect to a microcontroller through I2C serial communication and to an external sensor via a 11-pin digital input bus.