University of Cambridge > Talks.cam > Data Intensive Science Seminar Series > From parallel computing to parallel machine learning: Harness the power of parallelism for distributed optimization processes

From parallel computing to parallel machine learning: Harness the power of parallelism for distributed optimization processes

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Sri Aitken.

Parallel computing has transformed the way we approach computational problems by breaking down complex tasks into smaller ones that can be solved simultaneously across multiple processing units. Building on this paradigm, parallel machine learning extends the concept of parallel computing with a specific focus on optimizing multivariate objective functions in machine learning models. As AI applications continue to scale up and confront vast amounts of data and/or complex models, an imperative question arises: how do we train AI models that are bigger than the maximum capacity of a single computer? In this context, parallel machine learning emerges as a critical pathway for addressing these challenges efficiently. In this presentation, I will outline the key principles underlying parallel machine learning, highlighting its reliance not only on efficient divide-and-conquer strategies but also on effective collective communication among distributed computational nodes.

This talk is part of the Data Intensive Science Seminar Series series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

© 2006-2024 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity