Performance in Spark 2.0 and TBD

Agenda:

6:00: Food/drinks arrive

6:20: Talk #1: Performance in Spark 2.0

7:00: Questions

7:15: Talk #2: TBD

8:00: Questions

8:15: chill + relax = chillax

Description

Talk #1: Performance in Spark 2.0

Abstract:

Apache Spark 2.0 is the new major release of Spark, it has very interesting improvements and API changes. Brad will discuss what is coming up in Spark 2.0 how all of the components fit together to give better performance at every stage of the complete analytics cycle. He will also dive into the performance characteristics and the cloud/hardware demands of Spark SQL, Spark MLlib, and Spark GraphX. A preview of the work his group with the developer community to further improve Spark performance with hardware and software innovations will also be part of this talk.

Bio:

Brad brings over 30 years of experience in performance engineering at Oracle, Sun, Cray Research, and Floating Point Systems. Brad Carlile is Senior Director of Strategic Applications Engineering at Oracle, where he now has special focus on Analytics. He holds a bachelor's degree in engineering from Northwestern University and is the author of over two dozen technical papers in high-performance commercial and scientific computing.

Talk #2: TBD