I’ve just replaced MetroHash with CityHash and get 1000x (30ms instead of 30s) hashing speedup. It seems I somehow screwed MetroHash really bad. #cpp
I went ahead and benchmarked it. It is slower than us on all input sizes. It is also slower than other relatively fast hashes at small sizes, such as t1ha and xxhash3, although it may be higher quality than those, I don't know... more on that later :)
-
-
You might be interested in reading http://fastcompression.blogspot.com/2019/03/presenting-xxh3.html … both t1ha2 and mumv2 are latency-optimized (vs throughput optimized) hashes. XXH3 hash is inspired by them.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
They might show less throughput in throughput tests but has lower latency in intermixed load.
-
I've been talking latency numbers this entire time, not throughput. We actually don't do any throughput testing - all our measurements include a cpuid flush.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.