What is more CPU-intensive - an item or a tile?

[Tobruk]

If I dug the whole map - would it be worse or better than having a full-file world?

[SakuraKoi]

Just think about it...

Solid: Transfer energy

Gas: Transfer energy, move as whole, move mass

I think the answer is clear~

 

and yes, I can indeed confirm just how much of a difference it can make if you end up removing all tiles in a world for testing some other stuff. It's awful..

[Tobruk]

Just now, SakuraKoi said:

Just think about it...

Solid: Transfer energy

Gas: Transfer energy, move as whole, move mass

I think the answer is clear~

 

and yes, I can indeed confirm just how much of a difference it can make if you end up removing all tiles in a world for testing some other stuff. It's awful..

First you suggest that tiles are more CPU-intensive because of heat transfer, but then you say removing all tiles is "awful". 

[Saturnus]

7 minutes ago, Tobruk said:

First you suggest that tiles are more CPU-intensive because of heat transfer, but then you say removing all tiles is "awful". 

No. He did not. He in fact said the opposite. (To the first part.)

[Sasza22]

Heat transfer happens with both tiles and dug out items. But with tiles it`s pretty much all that happens. Dug out items are also calculated for the resurces available which requires them to be accessible to dupes and that`s pathing calculations as well. The reduced mass doesn`t matter since it`s the same amount of items (well maybe a bit less if same types of material stack). Not even mantioning the extra space you create that gets occupied by gas that gets the smae calculations for heat transfer the solid tiles got and movement on top of it.

Putting materials in dedicated containers makes them stack and reduces the amount of calculations slightly.

[Nightinggale]

The most CPU friendly is to dig out the entire map, then pump all the gasses out to leave a vacuum. That way there are neither gasses or solids.

[Junksteel]

1 hour ago, Nightinggale said:

The most CPU friendly is to dig out the entire map, then pump all the gasses out to leave a vacuum. That way there are neither gasses or solids.

I saw people saying vacuum is worse than solid tiles somehow. I think @JohnFrancis tested it out.

[Saturnus]

11 minutes ago, Junksteel said:

I saw people saying vacuum is worse than solid tiles somehow. I think @JohnFrancis tested it out.

In many ways vacuum is handled like a zero mass fluid in the game.

[Nightinggale]

does that mean that the game will be faster if I fill up unused space with sandstone tiles or similar?

[Tobruk]

1 hour ago, Sasza22 said:

Putting materials in dedicated containers makes them stack and reduces the amount of calculations slightly.

Is there a debug-related way of putting things into containers? It would be near impossible to do this the normal way taking into consideration how much I've dug up and how few dupes I have.

[Nightinggale]

17 minutes ago, Tobruk said:

Is there a debug-related way of putting things into containers? It would be near impossible to do this the normal way taking into consideration how much I've dug up and how few dupes I have.

Auto sweepers could come in handy for something like that. I wonder if putting materials into a sealed room will do the trick too.

[ozymandias_3]

7 hours ago, Saturnus said:

In many ways vacuum is handled like a zero mass fluid in the game.

so vacuum good or bad? sorry in advance for asking.

[JohnFrancis]

I filled a map with Vacuum and then compared it to one filled with tiles of coal, all the same temp. The coal tiles provided far better fps. 

[ozymandias_3]

okay, thanks. Tiles are better than gases as well I assume? 

[badgamer123]

so the thing is fill every spot of the map not need with some tiles?

[caffeinated21]

It would be nice if we could focus more on optimizing our bases than our FPS...

[avc15]

22 hours ago, Tobruk said:

First you suggest that tiles are more CPU-intensive because of heat transfer, but then you say removing all tiles is "awful". 

elaborating on this. Each tile with a gas:

• (1) exchanges mass with its 4 neighboring tiles, so we compare mass and transfer a small amount of it 4 times per tic per tile.
• (2) in each of those 4 mass transfers the game has to calculate the new temperature. It's at least 5 times recalculating temp per tile per tic, up to as many as 8, depending how they have it optimized.
• (3) Now do conductive heat transfer, at a minimum each tile conducts with its 4 surrounding tiles, but then also add one more per layer, or up to 9 more if there's a tempshift plate.

(4) Additionally, each debris created by digging represents two more heat transfer actions that take place per tic. One with the atmosphere, one with the tile it's standing on.

Solid tiles only exhibit behaviors 2 and 3 above, so solids take way less computer power than gases / liquids.

 

[natanstarke]

So... maybe...tempshift plates also causes lag? because it will make more heat transfers? Maybe gases should stop transfering heat at all? If we put tiles everywhere maybe its better if we put insulation or neutronium with debug or ceramic normal mode? im also guessing that the huge ammount of regolith in space causes problems because of temperatures too so sealing entire space seems necessary.

i will try to vaccumm a room and put all debris there, it will take a looong time...

Edit: yes tempshifts causes a lot of lag better avoid using them everywhere like i did in my main base for cool looking x-x