Oracle has officially launched MySQL HeatWave Lakehouse, a groundbreaking solution that allows customers to query data in object storage as rapidly as data stored within a database. This marks a significant milestone in the industry and provides users with enhanced capabilities for integrating and analyzing external data with their internal transactional data. MySQL HeatWave Lakehouse is compatible with various object store file formats, including CSV, Parquet, and files exported from other databases.
With this innovation, Oracle aims to provide users with quicker query processing, faster data loading, and simplified automation for querying data in object storage. The impressive capabilities of MySQL HeatWave Lakehouse are made possible by MySQL Autopilot, a feature that employs machine learning to enhance the execution plan of queries based on historical query execution. MySQL HeatWave Lakehouse on Oracle Cloud Infrastructure (OCI) operates with AMD EPYC processors, making it a highly efficient solution.
The AMD and MySQL HeatWave engineering teams have collaborated to optimize MySQL HeatWave for AMD EPYC processors, offering customers excellent price performance for business-critical workloads, including real-time analytics on vast datasets stored in object storage.
Oracle’s MySQL HeatWave Lakehouse has demonstrated its exceptional performance through benchmarks. In a 500 TB TPC-H benchmark, it outperformed leading competitors:
- 9 times faster than Amazon Redshift.
- 17 times faster than Snowflake.
- 17 times faster than Databricks.
- 36 times faster than Google BigQuery.
Notably, MySQL HeatWave Lakehouse’s performance for loading data from the object store is also impressive:
- 9 times faster than Amazon Redshift.
- 2 times faster than Snowflake.
- 6 times faster than Databricks.
- 8 times faster than Google BigQuery.
The solution’s unparalleled performance is attributed to its scale-out architecture, allowing massive parallelism for cluster provisioning, data loading, and query processing with up to 512 nodes. MySQL Autopilot also contributes by automating metadata creation for object files and adapting dynamically to the performance of the underlying object store.
MySQL HeatWave Lakehouse is a comprehensive cloud service, providing transaction processing, real-time analytics, machine learning, data lake querying, and machine learning-based automation within a single MySQL database service. This service is part of Oracle’s distributed cloud strategy and is available in various deployment options, enabling organizations to benefit from its capabilities no matter where they need to run their software.