Talk

Apache Arrow. In pursuit of speed

  • In Russian
Presentation pdf

Apache Arrow is a columnar data format and a framework that allows you to store data in a vectorised format, transfer the data between system components and process it at a high speed. Arrow implements Zero-Copy and No-Marshalling concepts. These two factors improve application performance significantly, which, in turn, made the format so popular in data processing systems. 

In the scope of the talk, we will speak about:

  • How Arrow helps in working with data
  • What hides beside Zero-Copy and No-Marshalling concepts
  • What developers go to reach an ultimate performance
  • #ipc
  • #jni
  • #marshalling
  • #offheap
  • #zerocopy

Speakers

Invited experts

Schedule