dcsimg
 

Ray Offers an Alternative to Spark

by Developer.com Staff

The new open source project could speed up machine learning.

University of California Berkeley's RISELab is working on a new open source project called Ray that could offer an alternative to Spark. Ray is a distributed, low-latency, real-time processing framework that aims to be faster than Spark.

Ray is based on Python and uses a Redis server to keep track of system state. It is meant to be used for machine learning use cases. It is currently a pre-alpha release.

View article

This article was originally published on Wednesday Apr 12th 2017
Home
Mobile Site | Full Site