+- Forums - Open Redstone Engineers (https://forum.openredstone.org)
+-- Forum: ORE General (https://forum.openredstone.org/forum-39.html)
+--- Forum: Build Discussion (https://forum.openredstone.org/forum-50.html)
+--- Thread: New measure of redstone logic (/thread-1640.html)
New measure of redstone logic - RandOMFG - 12-18-2013
Good evening, pearsons and appledaughters,
I have through several minutes of deep analysis and massive cirklejerking invented a new measure for redstone as we know it. This newfinding measure has been given the name "herps"(or singular "herp") and is a measure of, as properinglish put it, logical efficiency of a circuit or device.
The formula of the measure is as following: (maximum) logic-operations divided by total ticks required to execute said operations. For example: an RCA adder would require two sets of xors, two sets of and gates and a singular or gate, which is a total of 5. A normal RCA adder would require 6 ticks to perform all of these operations and from there, the equation is simple: logic-operations / ticks = 5 / 6 = 0.8333... herps.
This measure would not always make sense when comparing two circuits made for completely different purposes. for example an efficient bin->dec decoder would have a herp of possibly 8.
I have totally no idea whatsoever of what you would do with this except for possibly compare CPU'es of different architecture or ALU'es with different operations.
Anyways, tell me what you think and I think it would be awesome to use this measure as an internal thing or just for fun. If you have any ideas of what this could be used for please let me know.
RE: New measure of redstone logic - Iceglade - 12-18-2013
Parallel logic is a bit strange with this though, as you'll theoretically get a measure of the circuit's logical efficiency if all logic was done in series.
Or am I misunderstanding?
RE: New measure of redstone logic - AFtExploision - 12-18-2013
YUSHHHHH
RE: New measure of redstone logic - RandOMFG - 12-18-2013
That is the exact opposite of what it does. Something that calculates things in parallel will require less ticks and therefore have a higher herp.
This also gives me an idea. The measure could be used to determine a redstoners skillevel and a circuit's quality. If it has a higher herp than other circuits in the same category that would mean it either does more operations or the same operations in less time. With this we could make a player ranking based on average and/or total herp-level
This gives me another idea of considering size but I'm not sure how we would balance that. A new formula could for example be: operations/(ticks*qbrt(x*y*z)) though I reckon this would be extremely biassed.
RE: New measure of redstone logic - Iceglade - 12-19-2013
Aha, I see; so you wouldn't be able to compare the herps of an adder to the herps of a decoder. I see, very interesting.
and nah, that seems counter-intuitive; say you had a serial encoder sticking out the front of your device... and nobody likes cube roots anyway
EDIT: Or would you? Decoder would actually have a very high herps ((2^bits)/2 without repeater delay) due to low denominator, so quite possibly it would not be comparable for the exact opposite reason.
...I like math too much.
RE: New measure of redstone logic - Guy1234567890 - 12-19-2013
Counting "logical operations" is based on the frame of reference of the observer. If, instead, the number of ticks per net function (e.g. adding, multiplying, etc.) was considered, it would then be easy to compare one device to another. For example, if I made an ALU which could perform functions x,y,z in n ticks, then I would compare it to other ALUs which could do the same tasks and have a measure of relative performance. This, however, is exactly what we already do, and is why I do not think the herp is a useful unit. The herp, however interesting, is not useful for comparing devices.
More useful, in my opinion, is the idea of comparing tick growth as a function of throughput growth to parallelism growth. For example, If I had a device which could perform its task in 3 ticks, but gained 2 ticks every 8 bits which were added, I could write that as 3,2. 3 ticks plus 2 ticks per additional byte. This, in my opinion, is a more useful assessment of the logical efficiency of a device.
RE: New measure of redstone logic - properinglish - 12-19-2013
I think that there is a good bit of utility in this unit. Consider that different logic gets have different amounts of delay, for example OR vs. NOR vs. AND (0, 1, and 2 ticks respectively). We can build very efficient circuits by manipulating the logic to take advantage of the inherent efficiency of the gates we're using and by minimizing delay that is not required for a logical operation - the Herp describes that quality very well.
RE: New measure of redstone logic - properinglish - 12-19-2013
I should add that the circuit with a higher Herp value would not necessarily be better than a circuit with a lower herp value, but I do think it describes an interesting and useful quality.
RE: New measure of redstone logic - RandOMFG - 12-19-2013
I kinda agree with all of you. It does make sense to some degree in some cases but for the majority of comparisons it is not a reliable measure.
If you have any suggestions for modifications to the equation I would be very happy to hear.
And in response to Ice, I later realized (x+y+z)/3 gives the same number. To make size count less we could also change it up a bit and do 2-(x+y+z)^-1
Wait, no. make that 2-2/((x+y+z)/3)^2
Source: http://i0.kym-cdn.com/photos/images/original/000/234/739/fa5.jpg
RE: New measure of redstone logic - greatgamer34 - 12-19-2013
Honestly, since this can't be used to compare an ALU to a say GPU, I think it isn't completely necessary. As a red stone community we do compare devices, just without labeling it with a unit. Say for example, my cle is 5 ticks and newomaster's is 4 ticks, there both the same size, therefore his is better. Without using a unit we can see whose device is better than another's. If there was going to be a unit it should involve devices of similar purpose such as cle vs cla vs rca. It should be (LxWxH)for a single section,(amount of logic gates)of a single section/(ticks)of a single section(throughput)of a single section
So (LxWxH)(#of logic gates)/(ticks)(throughput)=herps
Idk if this would work though.