easyice commented on issue #12826: URL: https://github.com/apache/lucene/issues/12826#issuecomment-1822845516
Sorry for the late reply, I got lost in the JMH wrong loop for a while, Now I got the correct result, `memorySegmentReadGroupVInt` is faster than `byteBufferReadGroupVInt` in java 21 ``` Benchmark (size) Mode Cnt Score Error Units GroupVIntBenchmark.byteArrayReadGroupVInt 64 thrpt 5 11.518 ± 0.297 ops/us GroupVIntBenchmark.byteBufferReadGroupVInt 64 thrpt 5 5.271 ± 0.505 ops/us GroupVIntBenchmark.byteBufferReadVInt 64 thrpt 5 4.632 ± 1.122 ops/us GroupVIntBenchmark.memorySegmentReadGroupVInt 64 thrpt 5 10.991 ± 0.351 ops/us ``` And there is no performance regression in java 17 ``` Benchmark (size) Mode Cnt Score Error Units GroupVIntBenchmark.byteArrayReadGroupVInt 64 thrpt 5 10.661 ± 0.414 ops/us GroupVIntBenchmark.byteBufferReadGroupVInt 64 thrpt 5 7.179 ± 0.698 ops/us GroupVIntBenchmark.byteBufferReadVInt 64 thrpt 5 4.998 ± 0.370 ops/us GroupVIntBenchmark.memorySegmentReadGroupVInt 64 thrpt 5 7.646 ± 0.470 ops/us ``` the benchmark is using the latest code with https://github.com/apache/lucene/pull/12833 I'm running `luceneutil`, it will takes few hours. -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
