I promised I would have done some proper #benchmark for the new #hashtable multi producer multi consumer I am working on for #cachegrand and here they are!
197 million of unique inserts per second
235 million of updates per second
The #test #hardware was an EPYC 7502p with 32 cores and 64 hw threads.
The benchmark tool is under the benches folder in the repository https://lnkd.in/egqdeEuv (at least until I merge the branch :))
The new hashtable is between 2 and 2.5 times faster than the previous one!
Please keep in mind that these are the numbers of the hashtable, not of cachegrand when queried using the #Redis protocol, I still have to integrate it :)
Hopefully this weekend I will be able to import the functionalities I need to implement from the old hashtable to be able to fully integrate it!