In the process of playing, we will encounter this synthesis problem many times:The input products have a chance of not being consumed..
Here is a brief listing, including:Three kinds of hero fluid in the hero cycle,Unstable Gaia Soul,Medicine Steel Ingot,Nether Star,Light Permeation Domain，Cells in the biological line,dopamine,Cosmic ingot and super-solid super-energy silicon rock blackbody、ultra-stable self-healing adhesive made by a quantum manipulatorWait.
Of course, writing a template is simple. Directly calculating the approximate input consumption based on probability can solve the problem, or more directly, justIgnoring the probability consumption of the input is also a feasible method.
The following is a template written according to probability:
However, filling in the input according to probability this way encounters a problem, namelyThe actual parallelism of the machine will be compressed to a very low level.. The parallel count that seems able to execute the recipe multiplier at once, in fact, will only executeThe number of items for probability inputParallelism.
For example, in the above example, the number of light-transmitting domains for making 40K is 10,000 times that of the original formula, so 10,000 parallel operations should be performed, but the actual parallelism is only the number of input photon-penetrating Ampro-coated glass pieces, i.e., 75 parallel.
To address this issue, the author proposes a feasible solution here, namely adopting a subnet input method to maximize the machine's parallel utilization.
First of all, understand that the culprit limiting parallelism is the input material being too low in quantity, so raisingthis part of the inputwill solve it.
Taking the production of the light-transmitting domain as an example, we add an input bus to the machine.Pull items that are consumed probabilistically.Then transfer the pattern to a pattern provider, attach it to the interface, and distribute it into the subnet. Meanwhile, on the subnet, use a storage bus attached to the interface of the main network, and mark the items that are consumed probabilistically.
HereMark the storage bus as write-only, raise priority, and return the items from the probability input to the main network(This operation is equivalent to returning the items in the template assembly to the main network.)
In the figure, the blue cable connects to the main network, and the pink cable connects to the distribution subnetwork.
For the main network part, the input consumes items with a probability; hereFill the pull quantity according to the desired machine parallelismThat's it, and also rememberdyeMain network inventory pull and subnet part (you don't want to cross-contaminate recipes and then dump to the subnet, right?).
If, like me, you use an inventory input bus here, items here may be taken by other crafting tasks, thereby reducing parallel efficiency, and may even cause stuck orders (that is, all of these items are taken by crafting tasks, leaving the current recipe short of raw materials).
Facing this problem, we can consider usingrenamingas a way to solve:
WillRename the automatically pulled probability-consuming material, separate it from the mainnet synthesis task to ensure normal operation.
Note that template writingThe input materials also need to be renamed, otherwise it will cause a shortage of renamed items.
It can be seen that the modified recipe can properly saturate the parallelism.
For other recipes, the same approach can also be used:
1. Magic line probability input formula:
A. Phantom Ionic Liquid
The input is Spirit Marrow Fluid, which can be directly extracted later (after all, it's used for making UHV capacitors; by this stage you've definitely been to the Elven Dimension, right?). Simply remove it from the template, and then use the stock input bus to pull the input.
B. The Three Hero Fluids of the Hero Cycle
Just compile according to EMI's probabilities, and give a bit more quantity of the perfect mythic gems pulled from inventory.
C. Medicine Steel Ingot
This should be the earliest probability input problem encountered; before using the Great Alchemy, this article's method is not recommended (there's no alternative).
Probability input is the essence of manipulation. It is recommended to pull more inventory, and at the same time, when using inventory input, you can consider renaming.
D. Unstable Soul of Gaia
The probability input is Gaia Crystal,Considering that Gaia Crystal has no other crafting recipes in the later stages, it can keep its name unchanged.
It's recommended to make more Gaia crystals, as the usage of this unstable Gaia Soul is really high.
E. Nether Star
For the template writing here, I did not modify it based on probability, because the probability itself is already very high, and the price of the undead rune is basically free.
Using the allocation structure described above can prevent the input undead runes from getting excessively stuck in the assembly.
No need to rename.
2. Probability input formulation for quantum manipulators:
Noticed the recipe for making Cosmic Ingot, there iscatalyst. The model of this scheme isthat cannot be written into virtual items, therefore, pleaseManually create this catalyst and put it into the subnet to pull.
Considering that the consumption probability is very low, the inventory pull quantity can be slightly less.
Dragon NanoandUlu Nano SwarmSimilarly, it has no high-consumption recipes, so you may consider not renaming it.
Cosmic SingularityOnly this one consumes the recipe; do not rename.
3. Probability input formula for the biological line section:
A. Stem cells and biological cells can be ignored (after all, can betwisted, i.e., the biological module of the nano-integrated processing center)
B. Frostfield Shards belong to an auto-increment recipe; the structure of this article does not apply.
C. Dragon cells may consider using this structure to handle Frostland fragments; you can also refer toPersonal thoughts on self-increment ordering for culture vat categories (stem cells, biological cells, Frostland fragments, etc.)This article.
D. Dopamine has two probability inputs.
In fact, you only need to consider perfluorobenzene; just have the brain directly input from stock (What, are you saying your biological extractor places orders?)
Perfluorobenzene is a fluid and cannot be renamed. It is recommended here to use an ME input bus to pull perfluorobenzene equal to your machine's parallel count *1B.
Perfluorobenzene has only two uses: one is dopamine, and the other is making a protonated fullerene screening matrix. The latter belongs to the francium production line. However, if francium powder is ordered, you can manually order a large amount to avoid frequent crafting in subsequent ordering tasks (it's really complicated).
At the same time, the amount of francium used is not large; making over ten thousand manually in one go is enough to last until the element replicator, and copying afterwards will suffice.
Therefore, if you place a large order for francium powder, you don't need to input it into the warehouse; you can just use the inventory input warehouse.
4. The formulation for neutron activation has already been demonstrated earlier, so it will not be repeated here.
If using a neutron activator, pay attention to your neutron accelerator's level and the machine's parallel count.
Personally recommend building the highest activator.
Summary:
For probabilistic input recipes, you can adopt the expectation-based writing and the subnet pulling scheme described in this article (the original template writing was too aggressive; enjoy this productivity aspect).
However, for recipes with particularly low consumption probability, it is highly not recommended to write them according to expectation and put them into the assembly (especially low parallelism, which is both slow and power-consuming).
The value pulled for the subnet should be comparable to the machine's parallelism. It can be slightly more to improve stability.
When using the inventory input bus, consider renaming. The pulled products should use the renamed items, and the template should also be filled with the renamed items.