[GitHub] [incubator-doris] Gabriel39 commented on a diff in pull request #9663: [Enhancement]Add prefetch in join node

[GitBox <gi...@apache.org>]

Gabriel39 commented on code in PR #9663:
URL: https://github.com/apache/incubator-doris/pull/9663#discussion_r877799221


##########
be/src/vec/exec/join/vhash_join_node.cpp:
##########
@@ -26,6 +26,8 @@
 #include "vec/exprs/vexpr_context.h"
 #include "vec/utils/util.hpp"
 
+#define PREFETCH_COUNT (size_t(32 * 1024) / CACHE_LINE_SIZE)

Review Comment:
   To achieve best performance, we should take more information about cache size.
   For example, l1d cache size is 64KB in AMD Epyc server processor[1] while 48KB in Intel Icelake one[2]. 
   
   [1] https://www.club386.com/amd-unleashes-epyc-7003-series-with-3d-v-cache
   [2] https://www.techpowerup.com/248825/intel-increases-l1d-and-l2-cache-sizes-with-ice-lake



-- 
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: commits-unsubscribe@doris.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org