b) a high temporal correlation ...between consecutive requests I assume you mean. If this is not about arguing that you can't test hash tables with integers and if you advice on how to construct a suitable set of lookup keys, that in your mind is typical, I will redo the benchmark with it as a test case. I will use a hit ratio of your liking as well. On Sun, Mar 16, 2014 at 11:57 PM, Mike Pall <mike-1403@xxxxxxxxxx> wrote: > Fredrik Widlund wrote: > > Benchmark of LuaJIT tables vs Java HashMap vs C++ unordered_map > > Sigh, the Lua program doesn't even use local variables in the > inner loop ... > > Oh, and before anyone thinks this really measures hash table > implementation quality: it's just measuring the overhead of branch > mispredictions and cache misses for an atypical use case of hash > tables. > > [Typical hash table accesses use a) string keys with b) a high > temporal correlation and c) they usually have a high hit-rate or a > high miss-rate, but it's rarely 50/50.] > > --Mike > >