Archive: 13 posts

---

I just want to know how much thermometer each logic gadget (basic and advanced logic) takes up. When I found out that microchips take up more space than logic gates, I suddenly realized what was creating so much lag in my calculator level. I had activatable microchips everywhere. I'd just like I list of how much thermometer each logic gadget takes up, if that is possible. I miss the good old pre-war LBP days when we didn't have to dodge all these stupid bombs.

2012-03-14 03:16:00 Author: Kalawishis Posts: 928

---

Some thing along the lines of an item thermo qauge for every thing would nice. Instead of blindly plopping things down only to find out you suddenly have an over heated level. I just emit the *iss out of every thing that I can.

2012-03-14 03:34:00 Author: Butaneflame Posts: 64

---

The change isn't really that huge of an issue. It would take a lot of components to bring it up to even the very bottom tick on the thermo.

2012-03-14 07:30:00 Author: Speedynutty68 Posts: 1614

---

Oh I have a lot of components... In total, I have roughly 48 microchips in my level which I don't need.

2012-03-14 13:18:00 Author: Kalawishis Posts: 928

---

Sometimes you can get rid of the need to use a microchip, just by using an AND gate. Unless there's something I'm not thinking of, instead of passing one node through a chip so an input can be disabled.... why not connect it to an AND gate instead? Also, related to the thermometer, I have noticed (through backing up my levels) that the disk size of the levels does not directly relate to thermo.

2012-03-14 15:37:00 Author: Ali_Star Posts: 4085

---

Different components take up different amounts of thermo. I only did a test with selectors and counters to see which was more efficient for storing data and 100 selectors took up 1/3 more thermo than 800 counters, which is a huge difference. If I remember right the selectors alone filled around 20% of the total thermometer so they're pretty expensive.

2012-03-14 16:08:00 Author: Ayneh Posts: 2454

---

Different components take up different amounts of thermo. I only did a test with selectors and counters to see which was more efficient for storing data and 100 selectors took up 1/3 more thermo than 800 counters, which is a huge difference. If I remember right the selectors alone filled around 20% of the total thermometer so they're pretty expensive. Wow! That is a huge difference. Especially when you consider you have 8 times as many counters. Does the number of inputs/outputs on a selector have any major effect? From what I've noticed, the attracto tweaker has a noticable effect on thermo when you place it for the first time.

2012-03-14 16:40:00 Author: Ali_Star Posts: 4085

---

I really have issues whith a chip that is a sensor for the change of a signal. So i have a dyno and can check how fast the car makes the wheel accelerate. The problem is that this chip eats about 1/3 of the thermo. Im sure thats because i have about 200 components on each sequencer (100 other sequencers, 100 batteries). So im sure that extra logic on sequencers eat much more than just logic on a normal chip. So you better watch to place as much as possible NOT on sequencers

2012-03-14 17:49:00 Author: aleparad Posts: 28

---

Instead of defining a strict rule "Don't use this or that because this is more expensive than that", I think one best focus on efficient methods to solve ones logic problem; if solution B has less lag, uses less thermo, is more reliable, easier to comprehend and extend than solution A, then yes it is time to replace solution A for solution B. If thermo usage and performance are pressing matters in your level then you should dedicate your time in finding a better solution. I have been working over 9 months on my level now, mostly on improving existing solutions, in order to lower thermo, squash bugs, increase performance, smoothness, reliability, etcetera and bring those to an acceptable level. If you make something so complex that that feature alone takes up a third of your thermometer then put all else on hold because obviously you have a major issue to fix that will be decisive for whether you should continue building your level.

2012-03-14 18:18:00 Author: Antikris Posts: 1340

---

Wow! That is a huge difference. Especially when you consider you have 8 times as many counters. Does the number of inputs/outputs on a selector have any major effect? It seems the number of inputs does have an effect. 100 2-port selectors take up ~1/10 of a single bar 100 100-port selectors take up ~2/3 of a single bar You need around 600 counters to reach 2/3 of a single bar This differs from what I remember. Instead of defining a strict rule "Don't use this or that because this is more expensive than that", I think one best focus on efficient methods to solve ones logic problem; if solution B has less lag, uses less thermo, is more reliable, easier to comprehend and extend than solution A, then yes it is time to replace solution A for solution B. If thermo usage and performance are pressing matters in your level then you should dedicate your time in finding a better solution. I have been working over 9 months on my level now, mostly on improving existing solutions, in order to lower thermo, squash bugs, increase performance, smoothness, reliability, etcetera and bring those to an acceptable level. If you make something so complex that that feature alone takes up a third of your thermometer then put all else on hold because obviously you have a major issue to fix that will be decisive for whether you should continue building your level. Well, if you were trying to make a kilobyte of memory it'll use over 8000 components no matter what, so knowing which component is cheaper is necessary.

2012-03-14 18:44:00 Author: Ayneh Posts: 2454

---

Hmm... interesting... But i just try to find the most simple & easy way to make my logic work. IMO if you feel the logic is to complex and you have a hard time editing it easily. i'd say it's about time you found a more simple way to get what you need working. of course that also will mean less thermo used in the level. but then again it really depends on your level type. *mew

2012-03-14 18:57:00 Author: Lord-Dreamerz Posts: 4261

---

Well, if you were trying to make a kilobyte of memory it'll use over 8000 components no matter what, so knowing which component is cheaper is necessary. I catch your drift. That actually is an example of optimizing a solution. But I doubt a sequencer with 200 components on it is the best way to measure acceleration.

2012-03-14 19:46:00 Author: Antikris Posts: 1340

---

I just want to know how much thermometer each logic gadget (basic and advanced logic) takes up. I don't think anyone's ever produced a comprehensive list, so you'll probably have to do your own tests. A few things to bear in mind with the thermo... Cumulative thermo costs are hard to predict, i.e. if when your level only contains object 'A' it uses one notch, and only contains object 'B' it uses one notch, then it won't necessarily use two notches when it contains both objects - it's usually much lower. Thermo costs don't usually scale linearly, i.e. 1000 instances of a component won't necessarily use 10 times the thermo of 100, so it's better to test within an order of magnitude of the number you think you're likely to need. Hidden components use much less thermo than visible ones. Most of the CPU-complexity for components is in the rendering, not the simulation, so 1000 selectors inside a microchip uses much less thermo when the microchip circuitboard is closed. I only did a test with selectors and counters to see which was more efficient for storing data and 100 selectors took up 1/3 more thermo than 800 counters, which is a huge difference. For your 1kB memory example, I tested hidden components within an order of magnitude, and got... 10,000 counters = 5.1 notches 10,000 two-port selectors = 5.9 notches ...which is not particularly significant, but I guess it's still worth using counters if you have no use for the additional "not Q" output from the selector. I'd also pondered whether it was possible to leverage the fact that, say, a 16-port selector could effectively store four bits of data, using less thermo than four two-port selectors, but I suspect the encoding/decoding logic would destroy any possible thermo savings.

2012-03-15 01:53:00 Author: Aya042 Posts: 2870

Posts: 1077139    Threads: 69970    Members: 9661    Archive-Date: 2019-01-19

Aus dem Archiv wurden alle persönlichen Daten wie Name, Anschrift, Email etc. - aber auch sämtliche Inhalte wie z.B. persönliche Nachrichten - entfernt.
Die Nutzung dieser Webseite erfolgt ohne Speicherung personenbezogener Daten. Es werden keinerlei Cookies, Logs, 3rd-Party-Plugins etc. verwendet.