Author Topic: SPH (Read 12110 times)

UB VVG · « **Reply #1 on:** January 24, 2020, 07:40:52 AM »

SPH can make the simulation of the cosmic sand table more realistic.
Before that, I would also like to know when this update will be released.

UB VVG · « **Reply #2 on:** January 24, 2020, 08:00:40 AM »

https://www.youtube.com/watch?v=bU1QPtOZQZU It would be great if the universe sandbox could make SPH do the same thing as this video.
On the basis of this video, add another diffusion wave, and the clouds in the atmosphere are blown apart by the shock wave, which is like being pulled apart, and can also reveal the mantle in the collision.

Lord DC · « **Reply #3 on:** January 25, 2020, 05:12:57 AM »

have some patience, and if you looked at the previews i imagine with enough user-end fine tuning it can look like that
(though your computer will plot your death)

Cesare · « **Reply #4 on:** January 25, 2020, 03:36:51 PM »

SPH would also help to be able to save craters. Lets hope that surface grids are further improved so that they have more features and bug fixes over time.

False Anomaly · « **Reply #5 on:** January 28, 2020, 10:22:57 AM »

Quote from: Cesare on January 25, 2020, 03:36:51 PM

SPH would also help to be able to save craters. Lets hope that surface grids are further improved so that they have more features and bug fixes over time.

I'm not sure that SPH would be able to simulate craters. The particles would have to be very small and plentiful for that, and that would probably make it impossible to run on a home computer.

Cesare · « **Reply #6 on:** January 28, 2020, 11:01:46 AM »

SPH particles would have to be the size of grains of sand or smaller for the SPH feature to work properly.

Physics_Hacker · « **Reply #7 on:** January 28, 2020, 07:50:23 PM »

Quote from: Cesare on January 28, 2020, 11:01:46 AM

SPH particles would have to be the size of grains of sand or smaller for the SPH feature to work properly.

If they made them at small literally any computer would just freeze up and probably crash. That would require literally trillions upon trillion of particles...no computer could handle that, dude.

Cesare · « **Reply #8 on:** January 29, 2020, 12:24:22 AM »

Well, developers could reduce the amount of memory each particle takes. If each particle used only 1 byte of RAM, it would be possible to have particles as small as grains of sand.

UB VVG · « **Reply #9 on:** January 29, 2020, 01:30:26 AM »

https://www.youtube.com/watch?v=PnhflL7-I3I The effect of this video is similar to that of SPH, but SPH is likely to reduce the quality of the Universe Sandbox

UB VVG · « **Reply #10 on:** January 29, 2020, 01:38:34 AM »

https://www.youtube.com/watch?v=HMcyKhL_t8I
https://www.youtube.com/watch?v=dajIf1BHq48
In the second video, there is a surface rendering in 1 minute, but I think it is very rough.

UB VVG · « **Reply #11 on:** January 29, 2020, 01:46:57 AM »

SPH can also be used for galactic collisions. You can take a look at my previous posts about galactic collisions.
http://universesandbox.com/forum/index.php/topic,61946.0.html in which I like the galaxy collision in the fourth video very much.
It is hoped that developers can apply SPH to every aspect of the Universe Sandbox.

Lord DC · « **Reply #12 on:** January 29, 2020, 03:15:41 AM »

It would be pretty par for the course for them to allow you to manually set how many particles a planet gets, with a warning for how absolutely machine-meltingly complex it can get

Cesare · « **Reply #13 on:** January 29, 2020, 11:56:58 AM »

I know that SPH can be used for galaxy collisions and formation.

How will the performance of Universe Sandbox handle with SPH? Will it get better or worse?

If SPH makes Universe Sandbox's performance worse, I suggest that developers cancel the SPH development.

False Anomaly · « **Reply #14 on:** January 29, 2020, 01:24:48 PM »

Quote from: Cesare on January 29, 2020, 12:24:22 AM

Well, developers could reduce the amount of memory each particle takes. If each particle used only 1 byte of RAM, it would be possible to have particles as small as grains of sand.

Abra kadabra, magically make these particles memory-efficient! 1 byte is completely unrealistic. I'm no programmer, but you can't just store temperature, viscosity, velocity, and so forth in one byte. Try storing the properties of a black hole in one byte.

Cesare · « **Reply #15 on:** January 29, 2020, 01:48:54 PM »

Maybe one day it may be possible to compress a lot of memory data in 1 byte of space.

Pizzaeater1K · « **Reply #16 on:** January 29, 2020, 03:05:35 PM »

It straight up isn't though. One byte of space can only represent one of 256 different things, an amount of data which simply isn't enough for a particle. Maybe a few hundred to a few KB is possible, but not one byte.

Cesare · « **Reply #17 on:** January 29, 2020, 09:29:54 PM »

Well, developers would have to figure out a super compression for memory so that a single particle does not waste computers memory. A few kilobytes is too much of wasted memory for so many millions or billions of particles.

Physics_Hacker · « **Reply #18 on:** January 30, 2020, 07:42:46 AM »

Quote from: Cesare on January 29, 2020, 09:29:54 PM

Well, developers would have to figure out a super compression for memory so that a single particle does not waste computers memory. A few kilobytes is too much of wasted memory for so many millions or billions of particles.

But that's the problem. You can only be so space efficient and still have the information you need to run the simulation and have it be interesting. You couldn't eve use one byte to store just the location, not even to mention all of the other important properties particles would need for a fully featured SPH implementation. They could get the space requirements very low, but not that low. There will always be a limit to how many particles you can simulate at a time. 100,000 ring particles already makes even the best commercial, custom computers stumble, and ring particles are pretty optimized. Because of the way they implement them I'm sure SPH particles will be even more optimized but you still won't be able to get enough particles to have them be any smaller than many miles in diameter if you don't want your computer to melt.

Cesare · « **Reply #19 on:** January 30, 2020, 02:52:10 PM »

I bet that particles physics would become a problem only because that they would use too much RAM. Like I mentioned before, if each particle used only 1 byte of RAM, then that would solve the problem, but if each particle uses 1 KB or more of RAM than that would be too much for computers to handle if they have 4 GB or less of RAM.

Particle physics as SPH would perhaps bring lots of bugs and break many existing features in Universe Sandbox, making its performance much worse than it is now, so developers would be to be well prepared before SPH works properly before releasing it. In my point of view, SPH is not an important feature, so developers can wait a few years before it is complete and that it works perfectly well in its beta testing phase before being release as an official update for Universe Sandbox.

UB VVG · « **Reply #20 on:** January 31, 2020, 08:40:46 PM »

It is hoped that the developer will send a beta version before then.

Cesare · « **Reply #21 on:** January 31, 2020, 09:12:12 PM »

When?

False Anomaly · « **Reply #22 on:** February 01, 2020, 04:39:29 AM »

Quote from: Cesare on January 30, 2020, 02:52:10 PM

I bet that particles physics would become a problem only because that they would use too much RAM. Like I mentioned before, if each particle used only 1 byte of RAM, then that would solve the problem, but if each particle uses 1 KB or more of RAM than that would be too much for computers to handle if they have 4 GB or less of RAM.

You don't understand. It's impossible to store a particle's properties in one byte. You just seem to think the developers can waggle a magic wand and make things like this happen. This isn't a case of waiting for technology to advance, you just can't store a SPH particle in one byte. It's unrealistic to think something like that will ever happen.

Craters don't even need SPH to work. We had craters before surface grids sorta broke them just by displacing the surface of a planet, and it worked. I do hope the devs fix craters and the super low heat transfer I've mentioned in the support and bugs section.

Cesare · « **Reply #23 on:** February 01, 2020, 06:17:05 AM »

I know that it is impossible to store a single particle property in one byte. I know because you told me. I was just disputing how good it would be for computer performance if it was possible to store particle information in a single byte.

UB VVG · « **Reply #24 on:** February 28, 2020, 11:29:47 PM »

我想知道，开发人员大概在什么时候可以发布SPH。我十分的期待。

Cesare · « **Reply #25 on:** February 29, 2020, 07:17:25 AM »

What did you day? I don't understand Chinese or Japanese languages.

UB VVG · « **Reply #26 on:** March 04, 2020, 09:01:29 PM »

Can't your browser be translated?
It is strange for the machine to translate Chinese into English, so I use Chinese directly.

Cesare · « **Reply #27 on:** March 04, 2020, 09:04:36 PM »

There is no problem for my web browser to translate any languages. I am using the new Edge Chromium browser.

UB VVG · « **Reply #28 on:** March 04, 2020, 09:09:01 PM »

It should be these two days, the universe sandbox has a beta version on the steam platform. I thought it was about the update of SPH, and found no change. I tried the planetary collision, but there was no crater.
I don't know what's updated.

UB VVG · « **Reply #29 on:** March 04, 2020, 09:25:56 PM »

After the update, the temperature of all the planets is minus 270 degrees Celsius, and can not be changed.