> On Sun, 14 Oct 2018 at 06:19, Thomas Munro <thomas.mu...@enterprisedb.com> > wrote: > > Cache-oblivious hash joins cause a lot of TLB and cache misses. > ... > (There is another class of cache-aware hash join algorithms that partition > carefully up front to avoid them; that's not us.)
Just out of curiosity, can you please elaborate more on this part (with references)? I'm thinking about this topic for a while, and I'm wondering, if by another class you mean something like this [1], then even if it's not us today, are there any issues that prevent from experimenting in this area? [1]: https://www.cse.ust.hk/catalac/papers/coqp_tods08.pdf