Header Ads








Qualcomm Zeroth NPU - CPU that mimics the human brain and nervous system

Qualcomm has recently introduced its latest breed of CPU that mimics the human brain and nervous system – the Qualcomm Zeroth NPU (Neural Processing Unit).

The Zeroth NPU is Qualcomm’s attempt to make devices “smarter, anticipate our needs, and share our perception of the world so we can interact with them more naturally.” In essence, the Zeroth NPU is driven by brain-inspired computing unlike traditional CPUs that are based on speed.
Zeroth-powered devices will not only mimic human-like perception but also have the ability to learn how biological brains do. Not only it can follow code, it can also learn from user feedback as well as the environment.




By mimicking the nervous system and brain of a human, Qualcomm Zeroth processors are able to push forth embedded cognition driven by what the company calls brain inspired computing. This computing project is envisioned with three main goals.
1. Biologically Inspired Learning
2. Enable Devices To See and Perceive the World as Humans Do
3. Creation and definition of an Neural Processing Unit—NPU



An NPU is a new concept being initiated by Qualcomm Zeroth. The company hopes to create, define, and standardize a new kind of processing architecture here in a way that will allow them to live side-by-side in future processors for devices – on SoCs, that is. Using this technology, users will be able to tap into what Qualcomm calls “training the device for human-like interaction and behavior.”
Qualcomm hopes to make use of the world of neuroscientists with mathematical models of the human brain’s biological neuron behavior and push this sending, receiving, and processing of information to the Zeroth processing unit of the future. They’ll do this by replicating the electrical pulses (called “spikes”) made by neurons when a certain voltage threshold is reached.



In the video above you’re seeing a Qualcomm Zeroth processor in action. Placed in a robot, Qualcomm has allowed this machine to learn as it goes. This robot uses dopaminergic-based learning – something so simiple as positive reinforcement – to learn and grow in intelligence.

No comments

blogmytuts. Powered by Blogger.