Hello clever people, I’m having trouble dissecting a performance problem. MWE below. The computation in the 3rd benchmark is the composition of the computations in the first 2 benchmarks. I expect the 3rd benchmark to take about the total time taken by the first 2 benchmarks. But it’s 2x as long a…

[image] jocklawrie: I still don’t get why the 3rd benchmark is slower than the sum of the first two. The dynamic dispatch should happen in both the 1st and the 3rd benchmarks right? No, the dynamic dispatch is determining (at runtime) which compiled method of processcollection to call based o…

Please help: Poor performance, what am I missing?

General Usage

stevengj October 25, 2022, 1:36am 6

No, the dynamic dispatch is determining (at runtime) which compiled method of processcollection to call based on the type of d[k], and this only happens in the third benchmark.

Topic		Replies	Views
How to maintain performance despite type instability when you know the type? General Usage question	5	967	August 3, 2017
Dispatch time using `Val` is an order of magnitude slower than looking up with `Dict` Internals & Design	4	1070	May 5, 2017
Unexpected Allocations and performance loss from concrete classes Performance question	18	553	May 11, 2022
Speed of dispatch with heterogeneous functions and types Performance	10	958	July 16, 2019
Allocations reported by `@time` General Usage	4	577	June 7, 2019

Please help: Poor performance, what am I missing?

Related topics