• View Cart
  • Contact Us

The 2012 Challenge:

One Million neurons in a large pizza box

The Big Artificial Brain (BAB) project initiated in 2011 by CogniMem Technologies and the European Laboratory for Sensory Intelligence (ELSI) will demonstrate a truly parallel computing system with one Million silicon neurons connected together. The system will occupy 11"x11"x5" and be capable of recognizing one pattern of 256 bytes among 1 Million in 10 microseconds.

Bibliography and other links

In 1993, Guy Paillet, one of the founder of CogniMem Technologies Inc, presented the concept of a self organizing trainable parallel neural network chip to IBM and worked with a team at the IBM lab in Essonnes, France lead by Pascal Tanhoff. The outcome of this collaborative effort was an ASIC trademarked by IBM as the Zero Instruction Set Computer (ZISC) chip. Two generations of ZISC were released: ZISC36 with 36 neurons in 1993 and ZISC78 with 78 neurons in 1999. Unfortunately, IBM discontinued the manufacturing of the ZISC chip in 2001.

In 2007, Guy Paillet and Anne Menendez (also founder of CogniMem Technologies Inc) started the design of a replacement for the ZISC. The goal was to fit one thousand (1K) neurons per chip and add a few features to the architecture. OKI was contracted to manufacture the ASIC and the first batch of CM1K chips was produced in January 2008.

Like its predecessors, the CM1K uses patents applied jointly by Guy Paillet and IBM during the development of the ZISC chip. It is sold under the terms of a license with IBM.

DESCRIPTION
US
Improved neuron circuit architecture US-5717832
Daisy chain circuit for serial connection of neuron circuits US-5710869
Circuit for pre-charging a free neuron circuit US-5701397
Circuit for searching/sorting data in neural networks US-5740326