kwai / blaze

Blazing-fast query execution engine speaks Apache Spark language and has Arrow-DataFusion at its core.
Apache License 2.0
1.3k stars 121 forks source link

optimize hash joins #563

Closed richox closed 2 months ago

richox commented 2 months ago

for aggregates:

  1. improve byte arena performance.
  2. use unchecked slices in accumulators.

for hash joins:

  1. supports SIMD in join hash map.
  2. supports bulk search and data prefetching in join hash map.
  3. supports more hash join metrics.