XiangpengHao opened a new pull request, #5856: URL: https://github.com/apache/arrow-rs/pull/5856
# Rationale for this change This draft PR is not meant to be merged. Instead, it serves as a record for people who want to improve parquet metadata decoding performance. It demonstrates three hacks that can easily improve decoding time by 40%: (1) Use a better allocator (e.g., [mimalloc](https://crates.io/crates/mimalloc)) (2) Move [statistics](https://docs.rs/parquet/latest/parquet/file/statistics/enum.Statistics.html) to heap, i.e., reduce stack memory movement. (3) Use SIMD to accelerate decoding varint, as provided by [varint-simd](https://github.com/as-com/varint-simd) crate. -- 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]
