Starting Date:
Prerequisites: Interest in low level programming, I2C, SPI, and the Rust programming language.
Will results be assigned to University: No
We are building a low-power transputer (parallel computer with microcontrollers) to study neuromorphic computing and run inference in low power devices (low power as in low computing power and low energy consumption).
It is named the Incipient Neuromorphic Transputer Initiatiave in honour of Inti the Inca sun god. And because it is an incipient effort to build a small parallel computer (i.e.: a transputer).
The prototypes are built around the Raspberry Pi Pico taking advantage of the powerful communication primitives that its microcontroller (the RP2040) offers.
The objective is to connect Raspberry Pi Pico’s using their PIO communication hardware and build a small ad-hoc transputer that we can write software for. The objective is to explore the suitability of these devices to run inference on sparse machine learning models.
Note: this project is in collaboration with Anand Subramoney.