Details
Poster
Presenter(s)
![Takao Marukame Headshot](https://confcats-catavault.s3.amazonaws.com/CATAVault/ieeecass/master/files/styles/cc_user_photo/s3/user-pictures/15441.jpg?h=857caebf&itok=WBjCpuDx)
Display Name
Takao Marukame
- Affiliation
-
AffiliationToshiba Corporation
- Country
Abstract
An analog-to-digital mixed circuit with resistors for static neurons was implemented on a CMOS IC chip. The analog neurons with 1k synapses operate around 10 mW. The firing delay about 1 microsecond is caused by the patterns of inputs and synaptic weights. A recurrent connection directly from the output to the input generates an oscillation. Dynamic firing is observed, indicating a data-converter function from static to dynamic, in which a firing pattern can be tuned by randomizing synapses. One instance of such applications is reservoir computing, where non-volatile memristive devices were further considered for readout and learning with timing-dependent plasticity.