Kore Binary Format Optimizes Big Data for Analytical Workloads
TL;DR. Kore is a new open-source binary file format providing superior compression and query speeds for big data analytical workloads. - Kore demonstrates a 38% compression ratio, outperforming Parquet in efficiency. - It offers a 131x query speedup using column pruning and predicate pushdown. - Native Spark integration allows seamless reading and writing with PySpark and Spark SQL.
- Kore is an open-source binary file format for big data analytics.
- It achieves 38% compression, compared to 63% for Parquet.
- Query speeds are 131 times faster with column pruning and predicate pushdown.
- Kore includes native integration with Spark and PySpark.