-
dbis
-
Publications
-
2013
-
Main-Memory Hash Joins
PVLDB 2013 (E&A) Reviews
partition will be sufficient. 8. The paper spends only 2.5 pages (sec 7) on hardware oblivious vs. hardware conscious comparisons, while 3.5 pages (sec 5 and 6) on the results on the justifications for NPO [...] results of RJ and PR, neither of which are shown in Figure 7. I wonder why the results are not shown. D5. Section 6.1. The third paragraph puzzles me - it says (1) PRO is faster, (2) the (PRO?) results are [...] if a non-bucket chaining version of PRO was used there? D7. Typos and awkward sentences. Section 6.5 "must is". Section 7.2. The first paragraph. "On the AMD machine, performance is not as good as even …