tustvold commented on issue #5855: URL: https://github.com/apache/arrow-rs/issues/5855#issuecomment-2155031305
I'm curious how you intend to SIMD accelerate varint decoding, its a case study in being unfriendly to that... Regarding the linked flamegraph I'm not sure I'm seeing memory overheads dominating, there isn't much time being spent in allocation/memcpy routines, nor are we getting decode throughput at a level where I would expect memory throughput bottlenecks to be relevant <1Gpbs. My reading of that flamegraph is the inlined varint decoding is dominating the runtime -- 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]
