Amazon EMR Serverless Enhances Spark Performance with Larger Worker Configurations

AWS Big Data · 2026-07-29 · data

Amazon EMR Serverless has announced the introduction of a new worker configuration featuring 32 vCPUs and 244 GB of memory, aimed at optimizing performance for demanding Spark jobs. This enhancement is particularly beneficial for users running complex queries, as it has demonstrated an average of 29% faster execution times across 126 TPC-DS and TPC-H queries. Additionally, the larger worker setup has resulted in a 29% reduction in costs, especially for shuffle-heavy, multi-table join queries, which are often resource-intensive.

Why it matters for certification candidates

This development is significant for individuals pursuing certifications related to big data and cloud computing, such as AWS Certified Data Analytics and AWS Certified Solutions Architect. Understanding how to leverage EMR Serverless effectively can enhance practical skills needed for these certifications.

Original reporting: AWS Big Data