Current Max Power Block Efficiency

    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    So many of you who play this game and are seasoned may not need to read this as you have likely found the best layouts for your power block engines. However, just as many people likely, like me, have been unable to find anything useful as far as layouts for engines beyond the tutorials posted by beetlebear which, while helpful (really, if you're having trouble with making power blocks on your own, you should look it up if you haven't alraedy), didn't help me maximize the recharge rate for use in larger ships. They may also be slightly misleading to those new to the game who have yet to grasp the methodology of the game.

    This being said, I am not claiming to understand fully how the recharge rate is calculated, nor will I explain my hypotheses on how it is calculated, as the purpose of this post is purely to give the results of my countless experiments.

    Now to begin.

    For our purposes, let's say the volumn in which we are placing the power blocks is a 5 x 5 x 7 space. The five layouts I will outline are fictional (tutorial), solid, checkerboard, sponge, and capped checkerboard, terms that I will get into when I explain each one. In the cross sections I show, "x" will represent a power block while "o" will represent no power block (either empty space or another block altogether).

    NOTE: These layouts are intended for larger sized ships. Use of these in small and some medium sized ships (namely those that do not have much room inside for personel movement) may be unseemly, aesthetically displeasing, or even altogether impossible.

    Fictional (Tutorial):

    This is beetlebear's tutorial power block, which I am mentioning only for the sake of reference. To learn more about it please see the tutorial at http://www.youtube.com/watch?feature=player_embedded&v=YZ0qTeXjeT4. In the volume expressed above, the recharge rate is 2,585.7 e/sec over 15 power blocks, which is an efficiency of 122.42% of a single power block's recharge rate of 140.8 e/sec.
    NOTE: "Fictional" is a term my friend created to explain the large increse of recharge rate despite only placing one block. It refers to the fact that the three axes on which the blocks are placed create a fictional volume that is only physical on three edges, even though the fictional volume is being calculated into the recharge rate.

    Solid Power Block:

    The basic layout of a power block engine is a pure, solid block which, in cross section, looks like this:
    x x x x x
    x x x x x
    x x x x x
    x x x x x
    x x x x x
    This pattern continues through all 7 layers, and produces a recharge rate of 6,585.7 e/sec over 175 power blocks. This provides a recharge rate of about 37.63 e/sec for each block. This is extremely inefficient, as a single power block can provide 140.8 e/sec recharge rate on its own, cutting this layout's efficiency to a horrible 26.725% of each block's possible recharge rate.

    Checkerboard:

    This layout is similar to the solid layout. The key difference is that it uses half the power blocks by utilizing the following cross section throughout the 7 layers:
    x o x o x
    o x o x o
    x o x o x
    o x o x o
    x o x o x
    This layout can produce a decent 12,022.8 e/sec recharge rate over 91 power blocks, nearly doubling. Double the recharge rate may seem low if you're remembering the previous efficiency, however you're using half the blocks, so the recharge rate per block is about 132.12 e/sec, which yields a nice efficiency of 93.84%.

    Sponge:

    This is a variation on the checkerboard, however it utilizes no adjacent power blocks, contrary to what the previous two versions have done. The layout is made of alternating layers of the following two patterns:
    Layer A:
    x o x o x
    o x o x o
    x o x o x
    o x o x o
    x o x o x
    Layer B:
    o x o x o
    x o x o x
    o x o x o
    x o x o x
    o x o x o
    These layers in the pattern of ABABABA in the 7 layers produce a recharge rate of 12,393.9 e/sec over 88 power blocks, which is 140.84 e/sec for each block: an efficiency of 100.03% per block. This layout provides the best recharge rate per block, and is thusly more desirable when both resources and credits are low, as it is more budget friendly, though not by much.

    Capped Checkerboard:

    As the name would indicate, this is an evolution of the checkerboard layout. This layout contains three cross sections:
    Cap A:
    x x o x x
    x o x o x
    o x x x o
    x o x o x
    x x o x x
    Body:
    x o x o x
    o x o x o
    x o x o x
    o x o x o
    x o x o x
    Cap B:
    o x o x o
    x x o x x
    o o x o o
    x x o x x
    o x o x o
    This layout merges the checkerboard and fictional layout ideas. By compacting multiple fictional layouts into a checkerboard pattern which amplifies its effect, the power recharge rate in the above mention space is 12,907.8 e/sec over 95 power blocks, which translates into 135.87 e/sec per block, which is 96.5% per block. While this layout only provides the second best efficiency per block, it does provide the best recharge rate within a specific volume, and is much more desirable on larger ships, stealth or otherwise. However, as it requires a few more power blocks than the sponge, it is a little less budget efficient.

    edit:

    So, one of my friends I play with recently informed me of an improvement he made to the capped checkerboard. It changes Cap B to the following:
    o x x x o
    x o x o x
    x x o x x
    x o x o x
    o x x x o
    This is simply pattern of Cap A offset up two, which allows for a greater efficiency of the four fictional layouts on that side. This also allows for an easier extension of the two caps outwards. In essence, it's a repeated pattern of plus signs strewn accross both ends of the checkerboard layout.

    Results:

    Overall, the capped checkerboard layout is the best, though as stated at the beginning of this post, it should be reserved for larger ships, as the smallest possible layout is a 5 x 5 x 5 cube. If making a fighter or other small to medium sized ships focused on combat, the use of both fictional and sponge layouts will be the most volumetrically efficient.
     
    Joined
    Jun 26, 2013
    Messages
    452
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    I think forget to mention the most efficent way to produce energy is a straight line along the length of the vessel rather than any \'reactor\' system.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    That\'s a single line from the checkerboard pattern. And if you can only afford to have a 1 x 1 x whatever sized layout on your ship, that is the most efficient engine. Like I said towards the beginning, these are meant for medium to large size ships.
     
    Joined
    Jun 28, 2013
    Messages
    67
    Reaction score
    0
    Stretch the checkerboard outwards towards infinity and you win. It is the sum of many straight lines.



    Small ships benefit a little from actually understanding reactor mechanics and tinkering with engine designs, but for big ships the highest-efficiency pattern for both power and thrusters is always lots of parallel 1x1xn lines that don\'t touch.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    Unfortunately, what you\'re saying does not match up with everything that I\'ve tested. I\'m not trying to sound like a stuck up prick, but I\'m guessing you didn\'t even bother reading any of it. You could have at least looked at the volume I was using as my examples and the numbers that were being produced. And before you say that it doens\'t all increase the same amount, I will go ahead and tell you that you\'re right, they don\'t, but not for the reasons you\'re thinking. I\'ve done around a hundred of tests with dozens of different volumes. While the numbers did in fact change, it was always the same accending order of recharge rate: fictional, solid, checkerboard, sponge, capped checkerboard. And if my words still haven\'t convinced you, try experimenting with it for yourself. I simply posted the results that I found in the best manner that I could.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    Btw most of these tests took place last night and earlier today, so the numbers are correct as of v0.09378; build 20130709_211349.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    After testing a 5 x 5 x 5 cube of both the sponge and capped checkerboard, I found that the sponge is better, though only just. Further testing is needed, but it is likely that a cubic sponge is better than a cubic capped checkerboard. However, my earlier examples with rectangular prisms still hold true with the capped checkerboard being the best.
     
    Joined
    Jul 7, 2013
    Messages
    1
    Reaction score
    0
    I appreciate the time you put in.





    I have duplicated your reactor designs with the same results. I appreciate you showing us these test results. This post is about getting the most out of each block or the volume to work with.

    Nootau mentions a power block set up that does not come close to being greater than the effeciency per block of the \"sponge\" reactor, and I did test for the effeciency of Nootau\'s reactor system compared to the \"sponge\" reactor.

    TangentialThreat mentions that if you stretch the checkerboard design to infinity you will win, which is impossible for a sandbox game. TangentialThreat also mentions that the highest effeciency pattern is lots of 1x1xn lines that are not adjacent to eachother, but that pattern is reffered to as a \"checkerboard\" layout by xDark_Ace. Both of the tests from xDark_Ace and I show that the checkerboard layout is not the most effecient in controlled volumes.

    I say again that this post is to help players find the most effeciency per power block, as well as show the highest effeciency for volume. It is extremely helpfull for advanced players and fans of extremely efficacious ships.
     

    MrFURB

    Madman of the Girders
    Joined
    Jan 31, 2013
    Messages
    1,116
    Reaction score
    413
    It is important to note that a 5x5x5 reactor cube is not efficient for a ship that\'s not small, and simply multiplying the amount of cubes is not nearly as efficient as taking advantage of your ship\'s size to maximize your box dimension bonuses. It is possible to get over a million regen/sec with only a couple hundred blocks, where it would take much more If you just expanded 5x5x5 cubes.

    The cubes are only there to attempt to teach new people the ropes of how power works, and as far as I\'m concerned it is doing a good job.
     
    Joined
    Jun 27, 2013
    Messages
    202
    Reaction score
    10
    • Legacy Citizen 2
    • Purchased!
    so what\'s the best for large ships ? is it the capper checkerboard ? and if it is how do you make it larger, let\'s say 10x10 ? can you make it rectangular instead of cubic ? if it\'s true how a 5x10 would be ?
     
    Joined
    Jun 24, 2013
    Messages
    15
    Reaction score
    0
    I\'ve done around a hundred of tests with dozens of different volumes. While the numbers did in fact change, it was always the same accending order of recharge rate: fictional, solid, checkerboard, sponge, capped checkerboard. And if my words still haven\'t convinced you, try experimenting with it for yourself.


    A hundred test and yet you never bothered to use checkerboard lines with a lenght longer than 8?

    As soon as you use 8 or more in length, the lines (aka checkerboard) get a lot more effective than the sponge design..

    For example: 2x2x9: checkerboard 2559 reg; 2x2x9 sponge 2535

    and the longer they get the bigger (massively bigger) the difference gets.
     
    Joined
    Jun 28, 2013
    Messages
    67
    Reaction score
    0
    For all large ships the stretched checkerboard concept is superior. It scales and becomes more efficient with size while the other patterns don\'t. A typical high-end capital ship looks like a bundle of spaghetti when you cut it open because all the power, thrusters and probably also some of the weapons will be in 1x1xn lines.
     
    Joined
    Jul 2, 2013
    Messages
    4
    Reaction score
    0
    • Purchased!
    • Legacy Citizen 3
    The checkerboard pattern is also the easiest to build: Enable to symmetry planes, set length to maximum, and place them as fast as you can click. Or have two types of block, and switch between them, quickly filling a solid block of two checkerboard patterns.
     
    Joined
    Jun 24, 2013
    Messages
    518
    Reaction score
    0
    Having multiple lines is not as efficient as a single line.

    Two lines that are 25 long each (51 total blocks with core):

    10,957.4 e/sec

    One line that is 50 long (51 total blocks with core):

    16,039.0 e/sec
     
    Joined
    Jun 28, 2013
    Messages
    67
    Reaction score
    0
    There is rumored to be a soft cap on power gained per block around 50 blocks. Someone who is not lazy should test this.
     
    Joined
    Jun 24, 2013
    Messages
    518
    Reaction score
    0
    Rumor proven to be false:

    http://www.reddit.com/r/Starmade/comments/1hn9q6/info_starmade_by_the_numbers_part_2_power_blocks/
     
    Joined
    Jun 28, 2013
    Messages
    67
    Reaction score
    0
    Rumor proven to be false:

    http://www.reddit.com/r/Starmade/comments/1hn9q6/info_starmade_by_the_numbers_part_2_power_blocks/





    Well then. That series of Reddit posts should be a sticky.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    Again, that\'s not what my tests indicated. I\'m not saying your wrong, but just try replacing the two ends of the checkerboard with caps similar to mine and see if there\'s a significant difference. If you don\'t numbers to be significant enough for your liking, or you just like doing it your way, that\'s fine, but I\'d at least like to see some solid evidence that disproves my results instead of just disagreeing with what I said. If I\'m wrong, I\'m wrong, but if I am I would like to know so I can alter my engines to make them better.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    From my tests it\'s either the sponge or the capped checkerboard, however one of my friends told me that the checkerboard is better than sponge at much larger scales that I only touched on. If you\'re going for maximum recharge rate in a set volume without worrying about block count (block count only matters for thrust, cloaking, and jamming as far as I know), then capped checkerboard is the best. However, if you want to cloak, jam, or even both at the same time - especially while moving - then either the sponge or normal checkerboard is the best, depending on the size of the ship, as it provides the best average recharge rate on a per block basis.
    As for the dimensions, they can be whatever size you want, and it is probably best that you use dimensions that work best with your designs. Any of these layouts can be expanded by any number in any direction, so you can make a 5x5x5, 9x7x6, or even 99x99x100. And in order to make these larger designs, you simply make the first 5x5 pattern at whatever length you want, and then add layer by layer, simply copying the pattern of the two caps. That being said, one of my friend edited one of the caps to make it even more efficient. I will be adding an edit note to the capped checkerboard portion of my original post.
     
    Joined
    Jul 8, 2013
    Messages
    10
    Reaction score
    0
    • Legacy Citizen 2
    • Legacy Citizen
    First off, I\'d like to say that I do appreciate the feedback, and even more so that you included your own test results in that feedback.

    Now, what you said is true, but it only applies to ships at a certain scale range, but once you get that big you need a much larger recharge rate than 16,039.0 (that\'s assuming your ship doesn\'t not have a cross section of less than 10x10). If you make a stick ships that barely reaches 3 blocks away from that power core in any direction, then a single line would work perfectly fine. However if I want a fighter no longer than 20 blocks, then the difference between a single line and multiple lines is negledgable. Not to mention that doesn\'t even account for my capped checkerboard layout, which is better than the checkerboard layout (the multiple lines you were talking about). And what if I need a power core layout that contains 1,000 blocks for a capital ship? Are you telling me that I should make a ship over 1,000 blocks long to accomedate that power strip? Sorry, but I just don\'t see the efficiency in space or recharge rate at those scale ships. Large ships need huge power cores much larger in cross section than 1x1xn, and as I said power cores for medium to large ships are the focus of this post.