Overview
Learn how to achieve significant performance improvements and cost reductions in batch ETL workloads through GPU acceleration in this lightning talk from AWS re:Invent 2024. Discover the implementation of Apache Spark on Amazon EC2 G6 GPU instances, which delivers up to 5x speedups and 80% cost savings on Spark operations. Master techniques for integrating GPUs with Spark batch ETL workloads on Amazon EMR, and explore a specialized tool for identifying optimal workloads for GPU acceleration. Examine real-world success stories from organizations using RAPIDS Accelerator for Apache Spark, including detailed insights into customer adoption processes, production deployment strategies, and the substantial performance and cost benefits achieved. Presented by NVIDIA, this AWS Partner session provides practical guidance for leveraging GPU acceleration to enhance data processing efficiency and reduce operational costs.
Syllabus
AWS re:Invent 2024 - Accelerate Apache Spark up to 5 times on AWS with RAPIDS (ANT208)
Taught by
AWS Events