Smallest reliable production time

[eradicator]

To me recipe calculation includes inserters being woken up.

And to me it does not (i also stated that) because i think it depends on too many other factors.

On further thought what your setup seems to be measuring is the compressibility of inserter activity. So if the inserters already work in (bursts of) max-capacity transfers the impact should theoretically be much lower. For example if a recipe produces output bursts of exactly 12 items (==stackserter capacity). Or if the inserter workload is already compressed like for example if you compare a 10-fold-increased "0.5s recipe" with a 100-fold or 1000-fold increased version.

[Honktown]

On further thought what your setup seems to be measuring is the compressibility of inserter activity. So if the inserters already work in (bursts of) max-capacity transfers the impact should theoretically be much lower. For example if a recipe produces output bursts of exactly 12 items (==stackserter capacity). Or if the inserter workload is already compressed like for example if you compare a 10-fold-increased "0.5s recipe" with a 100-fold or 1000-fold increased version.

That makes me think, if I did a megabase-y bulk-recipe thing, it may be worth matching it to the smallest stack size among the ingredients * some n value - then each inserter can exactly insert the maximum amount needed (even if you have 1000 inserter bonus, an inserter can only grab 1 stack of an item at a time).

[Koub]

[Koub] I let the discussion go until now, but it's got quite off-topic, and going.
I suggest either recenter on OP's question or start a new discussion on the "bulk recipe and UPS optimizations for megabases". Both would be OK too