Youwei Wang has posted comments on this change. Change subject: IMPALA-2809: Improve ByteSwap with builtin function or SSSE3 or AVX2. ......................................................................
Patch Set 21: Hi Jim. If you don’t mind, I want to show you some performance data of the API StringVal StringFunctions::Reverse(FunctionContext* context, const StringVal& str) in string-functions-ir.cc before/after it utilizes the BitUtil::ByteSwap(SIMD) routine. The experiment setup is: Hardware: : Intel(R) Core(TM) i5-4460 [email protected] Testdata: lineitem.tbl file from TPC-H dbgen using scale=10; whose size is 7.3G. It has 59986052 rows. Testquery: select reverse(l_comment) from tpch1.lineitem; Experiment result: Using std::reverse_copy: Fetched 59986052 row(s) in 1214.63s Using BitUtil::ByteSwap(SIMD): Fetched 59986052 row(s) in 1188.57s We do get some performance improvement here considering console output and HDFS file reading takes most of the runtime. I am not sure whether such result is persuasive enough. Please feel free to tell me your idea about this. Thank you. -- To view, visit http://gerrit.cloudera.org:8080/3081 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I392ed5a8d5683f30f161282c228c1aedd7b648c1 Gerrit-PatchSet: 21 Gerrit-Project: Impala Gerrit-Branch: cdh5-trunk Gerrit-Owner: Youwei Wang <[email protected]> Gerrit-Reviewer: Alex Behm <[email protected]> Gerrit-Reviewer: Jim Apple <[email protected]> Gerrit-Reviewer: Marcel Kornacker <[email protected]> Gerrit-Reviewer: Mostafa Mokhtar <[email protected]> Gerrit-Reviewer: Tim Armstrong <[email protected]> Gerrit-Reviewer: Youwei Wang <[email protected]> Gerrit-HasComments: No
