Question about proliferators

[Devanor]

If I'm playing on infinite resources setting, is there ever a time I should set machines to productivity over speed (if they have the option) from proliferators?

[josmith7]

If you start running into CPU/GPU bottlenecks.

On very long production chains the multiplicative effect of getting 25% bonuses actually lets you use fewer buildings and belt than if you were using speed. (And far, far, less shipping of material)

Let's say you're making 150 small rockets per minute.
With +product you'd need:
796 miners*
329 assemblers Mk.III
157 plane smelters
42 quantum chemical plants
31 water pumps*

With +speed you'd need:
2741 miners*
399 assemblers Mk.III (+70)
247 plane smelters (+90)
72 water pumps*
64 quantum chemical plants (+22)

Those extra buildings would use up a bit more of your computer's performance. But all the extra logistics shipping you'd need for mining and moving all the extra resources would be a much larger impact.

Sure you've got infinite resources, so you're not going to run out, but your computer doesn't have infinite computing power :D. Scale up far enough and it'll start to run out; lwering your frame rate and game clock.

------
* actual numbers affected by VU research level

[Devanor]

Originally posted by josmith7:

On very long production chains the multiplicative effect of getting 25% bonuses actually lets you use fewer buildings and belt than if you were using speed. (And far, far, less shipping of material)

Is there a rule of thumb for how many machines in the chain there should be before I should use production instead of speed, or should I just play it safe and use production for most chains that use more than one machine?

[teron]

Typically the first thing you proliferate is the actual proliforator paint since it gives extra charges. Thus making the paint go further and proliforator paint cheaper.

I usually start with more expensive products down the chain, then work backwards.

[Borfoid]

I've never played on infinite resources, I only ever use +production because I have limited resources.

If I can get +25% all through a chain then that's a massive saving of effort on my part because I don't need to rework all the inputs as often.

I guess like the other guy said you can cut down on footprint and if you don't have a decent rig you might get in to situations where the game runs slow so production with less buildings makes things run faster.

If you aren't experiencing those issues though then it's better to use +production and use the normal ratio of buildings - in my opinion - because you have inherent redundancy in your system.

Something goes wrong, you've got +25% ... If you depend on that +25% and something goes wrong, it's going to go wrong a lot worse than if your default position is 100% throughput on the chain without bonuses.

I've never really played a factory game before though - so I could be wrong.

[Rekal]

Originally posted by Devanor:

Is there a rule of thumb for how many machines in the chain there should be before I should use production instead of speed, or should I just play it safe and use production for most chains that use more than one machine?

Simplest rule of thumb is don't use the Speed proliferation for anything and use the Extra Product option on anything that allows it. The only thing Speed accomplishes is reduced footprint in exchange for higher power cost. The trade off is not worth it, and it cost you the materials for the proliferator spray too. There are hundreds of planets to build on, don't try to cram massive amounts of everything in one place.

Some will argue that Speed can reduce the number of buildings at certain stages more than Extra Product and that can help with keeping the factories frame rate up. I disagree. 25% Extra Product reduces the previous step demand by 20% all the way back each step to the miners when used on the smelters. That's 20% compounded a number of times equal to the production chain length. That means fewer miners are active, fewer logistic drones are feeding your miner's ILS, fewer logistic vessels are in flight to feed your factory. All of which compares favorably to the Speed option's fewer buildings claim.

[Marakith]

Think the one exception I've found for "speed" use is proliferating Rail ejectors and Rocket launching Silos, allowing you to launch more rockets/solar sails faster with a smaller footprint. Sphere layers take forever to build otherwise.

[Bobucles]

Productivity is faster than speed for most items. The value of productivity depends on the total value of the recipe. High end recipes like research cubes gain a HUGE benefit, it's effectively the same as creating 25% more factory out of thin air. This remains true until all but the lowest tier ores and bars, which are up to preference.

Currently, the biggest CPU sink is the sphere itself. It's a bit strange if you think about it, a simple geometric shape defined by less than a thousand vertices and flat surfaces should not stress a 20 year old nintendo. But that's an optimization problem for the devs to figure out.

[Kyrros]

Originally posted by Bobucles:

Currently, the biggest CPU sink is the sphere itself. It's a bit strange if you think about it, a simple geometric shape defined by less than a thousand vertices and flat surfaces should not stress a 20 year old nintendo. But that's an optimization problem for the devs to figure out.

Well, it's not necessarily the vertices, it's also that just about every piece of detail between each vertice while being constructed is also mapped... multiplying it by 100s (or 1000s?). If the devs were to remove the 'randomness' from things like construction order and construction progress 'detail' - especially when using sail nodes!. But then we lose the ability to build more customized spheres visually, and the Devs are very much in favor of player customization in this game. It's an unfortunate tradeoff for some and a blessing for others.

[pemmons1]

I recently upgraded to level 2 (green) proliferation. Because producing green proliferators requires continued production of yellow (even more actually), I went around and retrieved all the yellow stuff in the pipeline, including sucking it up from the spraying machines' "reserve", and deposited it into a single warehouse from which the producer of green could be fed. It was fun doing this and watching the belts and machines fill up with the improved version. The transition, and the benefit from it, would otherwise take a long time otherwise.

A few questions, though:

1) I know that to benefit from proliferators, all inputs to a machine must have been treated. If any one is not, there is no improvement to production. But what if one has been enhanced with green and the other only with yellow? Do you get the level 1 benefit or, again, no benefit at all because the level of the inputs is not matched?

2) To what extent does it make sense to take stored yellow-enhanced products and run them through proliferation again to give them the two little triangles? I'm guessing that it would be worthwhile for the more advanced products, or those made from scarce materials, but not, for example, ingots or circuit boards.

3) Is there any benefit from proliferation on materials used for production by the mecha's own replicator?

[Borfoid]

Originally posted by pemmons1:

1) I know that to benefit from proliferators, all inputs to a machine must have been treated. If any one is not, there is no improvement to production. But what if one has been enhanced with green and the other only with yellow? Do you get the level 1 benefit or, again, no benefit at all because the level of the inputs is not matched?

It will round down, you'll lose anything higher than the lowest bonus... I think.

Originally posted by pemmons1:

2) To what extent does it make sense to take stored yellow-enhanced products and run them through proliferation again to give them the two little triangles? I'm guessing that it would be worthwhile for the more advanced products, or those made from scarce materials, but not, for example, ingots or circuit boards.

Storing proliferated products is a bit quirky, at least from what I've observed. The game wont keep track of every individual resources bonus level so creating a store of items that have been sprayed and then introducing items of a different bonus level can taint the whole batch. At least that's what it looks like.

I try to keep everything normal, without being sprayed until it goes in to a production line. then I spray it before it's used - that way I don't have to worry about that being a thing.

Originally posted by pemmons1:

3) Is there any benefit from proliferation on materials used for production by the mecha's own replicator?

No.

[pemmons1]

>I try to keep everything normal, without being sprayed until it goes in to a production line. then I spray it before it's used - that way I don't have to worry about that being a thing.

But everything except buildings goes into a production line, sooner or later. When I began this play-through I resolved to exploit proliferation to the fullest, as an exercise and learning experience if nothing else. At first I wondered whether it was realistically worth the trouble just at the yellow level, but now I'd answer with a resounding yes, To postpone implementing proliferation at all means far more rippiing up and rebuilding if and when you finally decide to do it. When I got around to going green, 95% of the work was already done. Just redo the area where the green stuff is produced and watch it percolate through the system already in place. Very satisfying. The way I'd address the issue next time of somewhat obsolete stock is not to overproduce immediately. Just set the limit of storage facilities way down until the product's' value and efficiency can be more enhanced. The Dutch Actuary advised this time and again. I should have paid more attention.

[teron]

Originally posted by pemmons1:

>I try to keep everything normal, without being sprayed until it goes in to a production line. then I spray it before it's used - that way I don't have to worry about that being a thing.

The exception to that rule is you want to proliforate your Proliferator Mk.III, since it gives the paint more charges. So for example on the mark 3 paint, painting it with mark 3 paint it goes from 60 --> 75 sprays per unit.

Fuel rods are another one you want to prepaint after production, since for non antimater it gives extra energy and output. For antimater it just increases how quickly the fuel is consumed meaning you need less artificial suns since each one will have a higher output.

[Rekal]

Originally posted by pemmons1:

2) To what extent does it make sense to take stored yellow-enhanced products and run them through proliferation again to give them the two little triangles? I'm guessing that it would be worthwhile for the more advanced products, or those made from scarce materials, but not, for example, ingots or circuit boards.

I think it would make more sense to update your factories to include proliferation on the inputs just before they're sent to the assemblers than worrying about what is sitting in storage. That way new product sent directly to the factory is sprayed before use and anything pulled from storage is sprayed again.

The spray coater is smart enough to recognize if something already has the level of proliferation it is applying so running through multiple spray coaters won't cost you extra proliferator.

That said, storage is bad unless it's a transport buffer.

Originally posted by pemmons1:

3) Is there any benefit from proliferation on materials used for production by the mecha's own replicator?

Proliferation for Speed or Extra Products is not used in the Icarus replicator. The Icarus fuel chamber on the other hand -- proliferated fuel is definitely recommended. You'll get both the increased energy output and increased duration of burn from the proliferation, so always spray your fuel!

[pemmons1]

Originally posted by Rekal:

I think it would make more sense to update your factories to include proliferation on the inputs just before they're sent to the assemblers than worrying about what is sitting in storage. That way new product sent directly to the factory is sprayed before use and anything pulled from storage is sprayed again.

The spray coater is smart enough to recognize if something already has the level of proliferation it is applying so running through multiple spray coaters won't cost you extra proliferator.

It's an interesting idea which I will consider how to apply on new planets, Especially if it is feasible while keeping all the sprayers for a given area in one neat line close together. Otherwise the criss-crossing belts soon become unwieldy. Furthermore, except in the building mall, most outputs are used as inputs for further manufacturing. Usually there are two or more inputs for a single output. So given how many factories use iron ingots, for example, is it really ideal to spray them at every input point, rather than spraying them once and for all just after they are produced (so they are ready for use in the mall among other ways)? The fact that the coating is fed into the sprayers from elevated belts is both an advantage and a disadvantage. It helps a lot by allowing other belts to pass under them, but if I need to use a splitter on them I'm in trouble. You speak from much more experience than I have, probably. How do handle these logistics?

[Rekal]

Originally posted by pemmons1:

So given how many factories use iron ingots, for example, is it really ideal to spray them at every input point, rather than spraying them once and for all just after they are produced (so they are ready for use in the mall among other ways)?

Ideal and idealize are two different things. Namely something that is perfect, and something claimed to be perfect. I don't like to discuss using absolutes so I won't claim something as ideal. I aim for practical implementation and lowest cost failure states.

In this instance, the question is which is better; Use fewer spray coaters and the logistics involved by spraying the outputs of a process before it is split to all the various locations it can be used as inputs. Or having more sprayers and the logistics involved on each location's inputs.

My thought process is:
1) Practicality, spraying directly before the point of use means that you can specify exactly when and where you want to use proliferation. Also you can ensure that no mistakes on other not directly present builds can mess with the current factory.

2) Lowest Cost Failure, a power dip on your grid will cause the spray coater to miss sprays. If this happens on the input sprayers the loss is a couple of recipes cycles to fail to proliferate all at the same location. If this happens on the output sprayers now your storage is contaminated with unproliferated items, your ILS supply buffers get contaminated, your ILS demand buffers get contaminated, and all the different locations the items are sent as inputs are now at risk as well.

Honestly, nothings stopping you from going over the top and proliferating at both the output and the input of each step. If you're already proliferating the output, just add a few spray coaters on the inputs for that same process. The input coaters probably won't get used much, but they'll cover any misses just in case. The only real extra cost is the extra spray coaters and the belts involved.

Originally posted by pemmons1:

The fact that the coating is fed into the sprayers from elevated belts is both an advantage and a disadvantage. It helps a lot by allowing other belts to pass under them, but if I need to use a splitter on them I'm in trouble.

Maybe you are unaware of the elevated splitter options. Next time you play, select a spliter like you're going to construct it and then look at the right side of the screens. You'll see a bunch of shortcut keys pop up listed there, why would you need a rotate option for a 4-way splitter? Try them all, especially the one you don't recognize. You'll be pleasantly surprised.