Dear authors: Please make the minor revisions suggested by both referees, and then I believe that this paper will be ready for publication in GMD. Thank you!

The authors have done a good job addressing most of the issues raised by the referees, and I largely agree with their judgement in choosing to not make some suggested revisions. The only case that I have some disagreement with is their decision to not make any revisions to address the anonymous referee's comments about the superlinearity discussion in Section 6.1. I agree with the referee that this discussion would be strengthened by providing some additional measurements collected with hardware counters to demonstrate the authors' belief that the superlinear speedup is due to improved cache utilization. However, I do not think this is critical, and I respect the authors' decision to not do so. However, I think that the wording here should be amended to make it clear that the authors believe that the superlinear speedup is due to cache effects, but that this has not been verified. Where the authors state "We explain superlinearity by better cache usage", the wording may lead some readers to believe that this explanation has been verified either experimentally or using some cache model. I suggest changing this to something like "We believe the superlinear speedup is due to improved cache usage, but we have not investigated this". It may also be useful to incorporate some of the content they wrote in their justification of their response to the referee on this point, though I leave this to the authors' discretion.

Beyond addressing the above, very minor revision, I think this paper is ready for publication. I thank the authors for their efforts.

Thank you to the authors for addressing the reviewers' and my suggestions and for this contribution to GMD!