nvidia / spark-rapids-tools   24.12.0

Apache License 2.0 GitHub

User tools for Spark RAPIDS

Scala versions: 2.12

RAPIDS Accelerator for Apache Spark Tools

This repo provides the tools to use RAPIDS Accelerator for Apache Spark.

Catalog

  • RAPIDS core tools: Tools that help developers getting the most out of their Apache Spark applications without any code change:
    • Report acceleration potential of RAPIDS Accelerator for Apache Spark on a set of Spark applications.
    • Generate comprehensive profiling analysis for Apache Sparks executing on accelerated GPU instances. This information can be used to further tune and optimize the application.
  • spark-rapids-user-tools: A simple wrapper process around cloud service providers to run RAPIDS core tools across multiple cloud platforms. In addition, the output educates the users on the cost savings and acceleration potential of RAPIDS Accelerator for Apache Spark and makes recommendations to tune the application performance based on the cluster shape.