1. 13
  1.  

    1. 2

      without any noticeable impact on query latencies

      Wonder if the JIT is worth it if it seemingly had no impact on performance.

      1. 2

        Obviously this is biased because it’s generally the people with a bone who yap, but I don’t remember reading of anyone being happy with the jit, same at work, at best it does nothing and at worst it takes tens of times the query’s execution time to jit it and gain 10% perf compared to the interpreter. Maybe there are work loads out there which get massive wins, but they seem few and far between.