Daniel Lemire
Daniel Lemire
@EgorBo @danmosemsft Pre-haswell processors are still common, sure. It is not unlikely that the performance could be decent.
We now have support for pre-AVX cpus and for ARM NEON. Upcoming version of simdjson will support runtime dispatch. This may get resolved with the next version of simdjson. Upstream,...
Sounds good... would you issue a PR?
We use a heap, not a sorted array. Maybe you need a different data structure.
> Let's maybe construct a set of meaningful tests in a separate repo so we can re-produce results and lock down versions of the libraries? That's a good idea. >...
It seems that ``pluck`` was removed from the benchmark library but it is not clear what replaces it (``map`` supposedly, but it fails with an error on my system). So...
@Restuta > What do you think are good next actionable steps? If you do have time, it would be useful for you to create your own independent benchmark.
Sure, please issue a PR at your earliest convenience.
@richardstartin Do you agree to close this issue?