Neuromorphic Processing The potential to make smart technologies

Neuromorphic Processing The potential to make smart technologies actually smart

The Future is Here! Qualcomm’s Zeroth Platform

Why cognitive computing? Human brains are highly effective at certain tasks -Visual recognition -Categorizing and learning These tasks have proven difficult for traditional computers -Google finds a cat on the internet But what is it? Brain-inspired computational software architecture 3

Historical Context • 1945 -1958=Von Neumann and Harvard architectures are developed. • 1964 -present=Modified Harvard architecture has been implemented in various machines. • 1974 -present=Artificial intelligence and cognitive computation has begins to appear on the radars of several computer engineering firms. 4

• Heterogeneous Computing: System on a Chip How Zeroth Works • Deep Learning Algorithms • Biologically Inspired Networks • Artificial Neural Networks 5

Uses & Benefits Optical Character Recognition Visual Recognition Natural Language Processing Solving Real Time Contextual Problems Energy Efficiency 6

Future Implications ● ● ● Improved device vision Autonomous robots Machine learning New standards Big data Devices that can sense or perceive the world and collect data; understand the information collected; and act independently—all supported by the ability to learn and adapt over time. 7

Social Concerns ● ● ● Privacy Security Job displacement Unemployment Invasive sensing/tracking Experts say automation and intelligent digital agents will permeate vast areas of our work/personal lives by 2025. 8

@CCT 506 Neurochip neurochip. wordpress. com Susan Noh | Isaac Riddle | Yingxue Guo | Shaqueta Pierre