|
|
 Jason Dai presented at AI Field Day 2 |
This Presentation date is May 28, 2021 at 13:30-14:30.
Presenters: Arijit Bandyopadhyay, Jason Dai, Ramesh Peri, Roshan Dathathri
Follow on Twitter using the following hashtags or usernames: @IntelAI
Intel Graph Analytics & AI: An Efficient Way to Analyze Massive Datasets
Watch on Vimeo
Watch on YouTube
Graphs are playing a key role in big data analytics, providing insights in many domains through traditional graph algorithms and graph neural networks. As the size of these data sets increase, the computing power needed also increases and the software techniques to manage it becomes ever more critical. For example, there are Intel-based single node systems which can have up to 16TB of main memory with Optane DC PMM, large clusters of machines are available with thousands of cores, and specialized hardware for processing large scale graphs are being developed. Intel and Katana Graph are collaborating to produce an efficient and scalable graph analytics library that works across this wide variety of platforms.
Presented by Arijit Bandyopadhyay, CTO – Enterprise Analytics & AI and Head of Strategy – Cloud and Enterprise, Data Platforms Group, Intel Corporation, Ramesh Peri, Senior Principal Engineer at Intel Corporation, Intel’s Architecture, Graphics and Software Group, Intel Corporation, and Roshan Dathathri, Software Engineer, Katana Graph.
Personnel: Arijit Bandyopadhyay, Ramesh Peri, Roshan Dathathri
Simplifying Big Data AI with Intel Analytics Zoo
Watch on Vimeo
Watch on YouTube
Applying machine learning to distributed big data analytics plays a central role in today’s intelligent applications and systems. These problem settings have pushed the field to address issues of data scale that were almost inconceivable to AI researchers even a decade ago. To address these challenges, Intel has open sourced Analytics Zoo (https://github.com/intel-analytics/analytics-zoo), which allows users to build and productionize end-to-end Big Data AI pipelines on large-scale Xeon clusters. In this talk, we will show how Analytics Zoo help real-world users (e.g., Burger King, Ant Group, etc.) build end-to-end Big Data AI solutions on Intel Xeon servers. Presented by Jason Dai, Intel Fellow and Chief Architect of Big Data AI, Intel.
Personnel: Jason Dai