Visit UCR Return to Campus website - Take the COVID Screening Check survey

Breadcrumb

A Talk with Tajana Rosing

-
WCH 205/206

TITLE: Brain-Inspired Hyperdimensional Computing and Machine Learning Acceleration for IoT Applications

We live in a world where technological advances are continually creating more data than what we can cope with. With the emergence of the Internet of Things, devices will generate massive data streams demanding services that pose huge technical challenges due to limited device resources. Sending all the data to the cloud for processing is not scalable, cannot guarantee the real-time response, and is often not desirable due to privacy and security concerns. Much of IoT data processing will need to run at least partly on devices at the edge of the internet. However, running existing machine learning on traditional cores results in high energy consumption and slow processing speed. To achieve real-time performance with high energy efficiency, we need to rethink not only how we accelerate machine learning algorithms in hardware. In the first part of my talk I will discuss some strategies that have allowed our team to significantly accelerate commonly used machine learning algorithms. However, we also need to redesign the algorithms themselves using strategies that more closely model the ultimate efficient learning machine: the human brain. Hyperdimensional computing is one such strategy that is motivated by the observation that the human brain operates on high dimensional representations of data. This, in turn, enables robust and highly efficient implementation of most commonly used learning algorithms in both software and hardware. In the second part of my talk I will discuss the development of an efficient learning platform that leverages hyperdimensional computing models and the design of software and hardware architectures that are multiple orders of magnitude more energy efficient while being just as accurate as the state of the art.

Tajana Rosing .pdf (118.89 KB)
Type
Colloquia
Target Audience
Students
Admission
Free