Innetra T1 - Neromorphic Chip with Spiking Neural Network Accelerator
https://www.eetimes.com/innatera-productizes-snn-accelerator-as-neuromorphic-microcontroller/
Highlights:
- “The array looks a lot like an analog FPGA: You can think of it as being a sea of programmable neurons and synapses that can implement any sort of [spiking] neural network topology”
- This allows the T1 to analyze both spatial data (like images) and temporal data (like audio), as well as spatio-temporal data.
1 Reply
Nice. I learned about neural networks around 20 years ago, when I was 18. I experimented, implemented them in C++ and played around with pattern recognition. Later I used them to make some virtual creatures learn to survive in a small simulated environment. At the time I thought about the idea of trying it with spiking neural networks, but found very few information about them. Then my interest faded, when the complexity of my projects lead to training times of several days and I had not the money to bay better hardware. But still I am excited whenever I realize how far this topic has come nowadays. 🙂