Matei Zaharia: 2017 Designing Self-driving Networks Workshop


Wednesday, April 12, 2017
Location: McCaw Hall, Arrillaga Alumni Center

"Accelerating Real-Time Data Analytics with Weld"



One essential component for self-driving networks will be data analytics platforms that can react to network conditions in real time. Traditionally, writing such platforms has been a time-consuming process requiring expert programmers. I will describe ongoing research in the Weld project at Stanford, which is building a compiler and runtime for data-intensive applications that automatically converts code written using high-level libraries into highly efficient machine code for diverse hardware platforms. With Weld, developers can use high-level data analytics APIs such as Apache Spark Streaming, TensorFlow and NumPy and still achieve state-of-the-art performance in real-time applications.


Matei Zaharia Assistant Professor, Computer Science, Stanford University Matei Zaharia is Assistant Professor of Computer Science at Stanford, where he works on computer systems and big data. He is also co-founder and Chief Technologist of Databricks, the big data company commercializing Apache Spark. Prior to joining Stanford, he was an Assistant Professor of computer science at MIT. His recent projects include Apache Spark (a distributed data processing engine), Mesos (a resource manager for datacenters), and systems for machine learning, stream processing and privacy at scale. Matei's work was recognized through the 2014 ACM Doctoral Dissertation Award and the VMware Systems Research Award.