andygrove commented on PR #3845: URL: https://github.com/apache/datafusion-comet/pull/3845#issuecomment-4240198985
I ran benchmarks with different number of partitions. This confirms earlier discussions on the PR that immediate mode is beneficial with small number of output partitions. <img width="1500" height="900" alt="shuffle_throughput" src="https://github.com/user-attachments/assets/fc9b455f-60c4-44b7-bad7-22bc2d073aec"; /> <img width="1500" height="900" alt="shuffle_peak_memory" src="https://github.com/user-attachments/assets/f9677b8c-54a4-4376-90a1-d87de1e64384"; /> -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
