[0.17.60] Unpowered inserters still consume a lot of ups

[qegr]

Translated by google .

My factory uses power switch control to reduce energy consumption and computing consumption when production is stopped.
But the results are not ideal.

After troubleshooting, I found that the computational cost of the inserter did not decrease when it was not powered.

In order to verify this conjecture, I used the map editor to generate 80,000 inserters.
Below is the calculation cost and ups comparison between the power-on state and the unpowered state.

1.png (58.58 KiB) Viewed 3012 times

power-on state

2.png (161.61 KiB) Viewed 3012 times

unpowered state

3.png (160.56 KiB) Viewed 3012 times

Obviously the gap is not obvious.
I have not tested whether other entities are as also as inserters.
I don't know how the internals are calculated.But I think the computational cost of unpowered should be much lower than the computational cost of powered.
If it can be achieved, it is very practical to use power switch control in a large factory to reduce the computational cost. Hope can be achieved.

[Adamo]

This is actually a well-known effect of using power-switch controls, and gets brought up every now and then. I'm not familiar with the reasons why it is difficult to fix.

[Rseding91]

Thanks for the report however that's not a bug. Unpowered stuff has the same cost as powered stuff because there is no sleep logic for "has no power" to wake up things when they *do* have power. It
s expected that everything has power at all times.

[Trebor]

It s expected that everything has power at all times.

Then why give us power switches in the first place?

[Rseding91]

Because it sounded interesting to those that decide what features get put into the game and so it was added.

That's my guess anyway. I don't know the real answer.

[Deadlock989]

Then why give us power switches in the first place?

I doubt it was to save UPS.

Some machines drain power even when they have nothing to do. Gating them behind switched power allows you to isolate them, if that's what you need to do.

You can also use them to isolate power production, e.g. steam engines that only kick in and start eating fuel / producing pollution when solar and accumulators can't meet peak demand.

They're a game mechanic, not an optimisation.

[qegr]

In an older thread I proved powered down inserters wake up from sleep when items on a belt pass them and never go back to sleep until powered again.

If this is true, then I think that even if it is not a bug, it can be further optimized.

[Trebor]

In an older thread I proved powered down inserters wake up from sleep when items on a belt pass them and never go back to sleep until powered again.

If this is true, then I think that even if it is not a bug, it can be further optimized.

Someone suggested making sure that your belts were emptied or stopped before turning off power.

[Optera]

Someone suggested making sure that your belts were emptied or stopped before turning off power.

That someone was me in the same post.

In an older thread I proved powered down inserters wake up from sleep when items on a belt pass them and never go back to sleep until powered again.
The solutions to this are either to use filter inserters in whitelist mode without a set filter, those will sleep forever until they get a filter, or stop the items from moving past inserters, be it with stopping the belt itself or using a 2 stage power down that first cuts off resource input and once all machines are starved turn off power.