Accelerating Deep Learning

Speaker:  Michael Gschwind – Yorktown Heights, NY, United States
Topic(s):  Artificial Intelligence, Machine Learning, Computer Vision, Natural language processing

Abstract

As computing systems transform to become more responsive to human needs, Machine Learning and Deep Learning are poised to become key drivers of new cognitive computing systems.  Artificial Neural Networks hold out the promise of transforming entire fields of applications, subject to matching the available computing resources to this new application domain:  computing systems from yesterday's rule-based application architectures to tomorrow's cognitive holistic applications implies massive compute requirements to train and deploy artificial neural networks.  New, hardware-accelerated systems promise to meet these performance requirements and make the cognitive transformation a reality.  As an example, next generation systems optimized for cognitive computing based on numeric accelerators offer a sweet spot of scalability, performance and efficiency for Deep Learning applications.  This talk will present DL-optimized hardware systems, a review of key deep learning software architectures, deep learning ecosystem needs and an industry-first Deep Learning software distribution called PowerAI developed at IBM.  

As part of the focus on cognitive application enablement at IBM, the new server will be accompanied with a rich pre-optimized and prebuilt Deep Learning software distribution to simplify and accelerate  deployment of Deep Learning technologies with a simple to install software stack.

About this Lecture

Number of Slides:  40
Duration:  60-90 minutes
Languages Available:  English
Last Updated: 

Request this Lecture

To request this particular lecture, please complete this online form.

Request a Tour

To request a tour with this speaker, please complete this online form.

All requests will be sent to ACM headquarters for review.