Sources at Amazon Web Services tell us that the talk “Transforming Big Data with Spark and Shark” by Michael Franklin and Matei Zaharia received the highest ratings of all the regular (non-keynote) sessions at the 6000-person AWS developers/users conference held in Las Vegas last week.   The talk included a virtuoso live demo of both systems on a 20 node EC2 cluster by Matei, where he demonstrated order-of-magnitude speed ups over Hive, and also showed his ability to type error-free SQL and Scala at the command line in front of a large audience.  Very impressive!   The video has not been posted yet, but slides are available at:

http://www.slideshare.net/AmazonWebServices/bdt305-tranformingbigdata