Home icon

GPU-accelerated Apache Spark with Amazon EMR and NVIDIA RTX PRO 4500 on Amazon EC2 G7 instances runs up to 3.7x faster

Big Data Blog

This article announces GPU-accelerated Apache Spark performance benchmarks using Amazon EMR on EKS with NVIDIA RTX PRO 4500 Blackwell GPUs on EC2 G7 instances.

  • G7 instances complete TPC-DS 3TB benchmark in 4.7 minutes, 3.7x faster than comparable CPU instances
  • Cost per run is 31% lower on G7 despite 2.5x higher hourly rate due to dramatically reduced runtime
  • NVIDIA cuDF plugin automatically offloads Spark SQL operations to GPU with no code changes required
  • GPU acceleration excels on compute-heavy queries: 15 queries achieved 5x+ speedup, 102 of 103 queries faster on GPU
  • Ideal for AI/ML feature engineering, complex ETL pipelines, and real-time analytics workloads
  • Setup requires Amazon EMR 6.9.0+, EKS cluster with G7 nodes, and NVIDIA device plugin

GPU-accelerated Spark on Amazon EMR enables data teams to process larger datasets faster, iterate on models quicker, and reduce operational costs without modifying existing Spark code.

Go to article

The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.

Related articles

Jun 18
2026
Announcing Amazon EC2 G7 instances accelerated by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs
Sep 3
2026
Run Spark 31% faster and optimize compute costs with Amazon S3 Express One Zone on Amazon EMR
Sep 1
2026
Accelerate Apache Spark debugging on Amazon EMR with AWS DevOps Agent
Jul 29
2026
Accelerate Spark on EMR Serverless with larger workers and shuffle-optimized disks

The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.