Author: @propylaia
Feel free to add if there are omissions or errors.
Applicable to normal mode.
I. Foreword
Advanced hypercube can use "coordinate information card" to bind 20 blocks, and defaults to sequential mode, which just meets the orderly requirements of the assembly line.
Version 0.5.4 added the 'coordinate label gun', which is even more powerful, making it more convenient to bind blocks.
II. Recommended Construction of Assembly Line
After building the main block, I personally recommend using a wrench to right-click the "main block" to rotate and place the assembly line.
At the same time, the terminal sets the number of repeated structures to 17, which is the maximum size. This arrangement facilitates batch construction and does not take up too much space.
The hypercube and supply/interface can be placed at the top.
Note: Normal mode can use quadruple input bins, while expert mode requires four input bins.
3. Using the "coordinate label gun" to mark the input bus and input hatch.
Use gun from low to high "right-click" Mark the input bus of the assembly line and the quadruple input hatch
After all 17 compartments are marked, coordinate label gun "shift+left click“The hypercube can be imported into the hypercube.
Note: The backpack needs to carry 17 coordinate information cards. When there is an AE wireless terminal in the backpack, it will automatically use the coordinate information cards in the network.
4. Common Assembly Line Supplier Configuration
traditional method
Place the template supplier on top of the output assembly and enable its 'Lock Synthesis' feature to complete the setup.
Assuming that when mass production is needed, each assembly line uses the same construction method, and finally uses a memory card to copy the same template to achieve parallelism.
Use hypercube 'polling' mode with 'sequential' mode
Place the hypercube anywhere and use a screwdriver to switch the hypercube into polling mode. (as superior)
The hypercube on the assembly line remains unchanged. (as a subordinate)
The original template supplier moved to the edge of the hypercube of "superior",
Send the finished product back to the network with the ME interface attached above the output assembly.
Finally, in the "superior" template supplier, turn on "Pause sending when there is synthetic material in the container after parallel distribution" to complete the setup.
5. Advanced Assembly Line Supplier Configuration and Additional Notes
Change the blocking mode to "Block insertion when templates are not the same" to achieve full parallelism.
Also, because of the use of the "Giant Input Bus", if there are more than 64 items of materials in the recipe, the order will be messed up. The solution is to use the "Template (Named)" to intervene in the recipe.
For example, LuV Stance Generator
After modification, just rename and rewrite the template for the coil using the stamping machine.
6. Expert Mode Advanced Assembly Line Additional Settings
Due to the additional requirements for fluid order in expert mode, simply using a setup like the one shown below may cause a problem.
Taking the UV stage as an example, at this point 4 UV input buses are used. To fully utilize the parallel processing of the machine, usually the pattern provider needs to be set to 'Prevent insertion when not the same pattern'. Then take making a Wetware Processor Supercomputer as an example, one batch makes 10,000. However, the UV input bus capacity is 2048B. After the pattern provider distributes 2048B of solder to the first input bus, since the second and third input buses have PBI and Tin Plasma respectively, the pattern provider will continue to distribute solder to the fourth input bus, causing the fluid to be out of order. A simple solution is to change the pattern provider setting to 'Pause transmission when synthetic materials exist in the container after parallel distribution', but this means that parallelism is difficult to fully utilize (unless manually doubling the recipe) and bears the cost of multiple overclocking penalties.
The essence of the above problem is that the cache of the UV input hatch is still not large enough. We can place a super tank before the input hatch to buffer the fluid. However, the interaction between the hypercube and the super tank is not that intelligent. Looking at the Jade display of an empty input hatch and an empty super tank, we can find that when the super tank is empty, Jade displays "Air 0mB", while the empty input hatch displays "0mB". If we change the hypercube's original binding of 4 input hatches to binding these 4 super tanks, the pattern provider suspects that it can only send air into the super tanks, thus not sending materials.
The solution is to change the fluid to a subnet, and the hypercube binds the ME interface of this subnet.
Subnet storage bus attaches to Super Tank, need to set priority, from bottom to top: 0, -1, -2, -3 (bottom priority higher is fine). Need to enable 'Report non-interactable items'. Pattern provider still set 'Prevent insertion when not same pattern', so that you don't need to manually double the pattern and can fully utilize the parallelism of loading.