Finally, A Blazing Fast Fluid Simulator! 🌊
HTML-код
- Опубликовано: 5 май 2020
- ❤️ Check out Weights & Biases and sign up for a free demo here: www.wandb.com/...
The shown blog post is available here:
www.wandb.com/...
📝 The paper "Fast Fluid Simulations with Sparse Volumes on the GPU" and some code samples are available here:
- people.csail.m...
- www.researchga...
📸 Our Instagram page is available here: / twominutepapers
🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible:
Alex Haro, Alex Paden, Andrew Melnychuk, Angelos Evripiotis, Benji Rabhan, Bruno Mikuš, Bryan Learn, Christian Ahlin, Daniel Hasegan, Eric Haddad, Eric Martel, Javier Bustamante, Lorin Atzberger, Lukas Biewald, Marcin Dukaczewski, Michael Albrecht, Nader S., Owen Campbell-Moore, Rob Rowe, Robin Graham, Steef, Sunil Kim, Taras Bobrovytsky, Thomas Krcmar, Torsten Reil, Tybie Fitzhugh
More info if you would like to appear here: / twominutepapers
Meet and discuss your ideas with other Fellow Scholars on the Two Minute Papers Discord: discordapp.com...
Károly Zsolnai-Fehér's links:
Instagram: / twominutepapers
Twitter: / twominutepapers
Web: cg.tuwien.ac.a... Наука
"Dear fellow scholars..."
Yep, I'm most definitely a scholar. I spend most of my days scholaring with other scholars at the local scholartorium. I was just saying to my friend earlier today (also a scholar), wouldn't it be great if everyone were scholars? Anyway, great vid. Scholar out!
👌
XD
I went to school! That makes me a scholar right? *Right?* :(
dropped out and do not intend to study again, do I qualify because I'm a curious idiot still ?
@@sankhyohalder97 Maybe you have to go to a "scholl".
That reminds me I've forgotten to do my German exercises today!
"If no one watches them, so be it, I still love doing this"
You are too kind Bernhard. Thank you so much! 🙏
Agreed
Ai breakthroughs grab my attention more often than fluid simulation but little by little, Two Minute Papers has worn me down and now I get excited for those too.
@@TwoMinutePapers So are there any softwares actually using this technique youre talking about then? Very misleading title if you dont actually share where you got all of this from
594k ** now 3 days later ;D
1:38 "The first thing you are probably thinking about..." [shows tsunami in minecraft]
Me: This man is psychic
That’s gonna be fun in 2b2t
Imagine portal with constant tsunami, ohh I’m trembling just thinking about it.
O C E A N M A N
*_THE OLDEST ANARCHY SERVER ON MINECRAFT_*
Shut up, it is a tsunami in Lego.
"If no one watches them, so be it; I still love doing this".
That's a fantastic philosophy, and it really shines through in the quality of the work you produce and the clear passion you express for this quite niche field. I don't do CGI or complex maths, but there's something addictive about watching you do what you love.
Also the simulations are pretty.
what a time to be alive
but on another note; those ferrofluid simulations are amazing
Completely agreed - absolutely incredible work! If you wish to see more, check this out. - ruclips.net/video/goD36hVVl7M/видео.html
@@TwoMinutePapers thanks! I had a look and wow; the paper and work in that is crazy, great stuff.
Magnetofluid dynamics are a heck of a thing
Well...THE time to be alive would be when all the physics and AI simulations show are mainstream happening at real time on videogames and such
@@estebansingh9411 I've been working on one with powered ragdolls in wind, and water that's part of the game mechanics, not just an effect, for 10 years as an experient, we called Kontrol and updated weekly for a while. Stopped. a few years back because of too few appreciated the challenge of wind effects and physics made it too hard to travel, and we had spent tens of thousands already just so creature didn't walk like a drunk. It's in 2d because we can see inside workings, there are allot of firsts, 3d oceans I predict in 10 to 20 yrs studios will do it more like kontrol does. THings like surfing on wind usign just air drag and friction, were discovered and I never felt greater thrill in my career than to see these things emerge, I tuned and shaped surfboards all night to see how they performed In
stead of coding the behaviour, and the custom sparse air model I used is nothing like what can be done with modern pc or phone. I firmly believe all aaa games will use physics for all movement, include air and pressure with balancing robot controllers that can fall. Its a time when playing games will make u better at sports, not worse. I want to add fluids because they are so fascinating, especially vortices. In a year or so I hope to release an android update with vids on my channel. It's rough but playable now somewhat.
I can't wait until some of these techniques materialise into common software
Me too, imagine this in Houdini 🥺
Blender has updated to use particle based fluids but I'm not sure if that means the sim calculates on your GPU like this video talked about. You still need to establish a domain, but it can be "adaptive" and grow or shrink based on what the fluid does.
Me too. I've been waiting for a good opensource fluid simulator that I can use for my micro-hydroelectric projects.
@@recklessroges Would love to see this applied to aerodynamics. Wonder if it can be used in a dynamic system, like simulating airflow over a moving propeller and aircraft. Being able to visualize the nuances of the air flow and try different configurations for an aircraft quickly would be great.
@@hashemieada4846 there are new sparse solvers in H18.. still no support for sparse gpu sim, tho.. i think they have to redo microsolvers for that to happen..
Two Minute Papers: "If you see here, the fluid counter..."
Me, who knows absolutely nothing about graphics: "Oh, of course."
Everyone: Wow, that's so realistic and fast!
Me: hehe water go splish
I wanted the buildings to not be immovable objects.
@@recklessroges The Day After Tomorrow
@@DATECStudio because Károly said so
@@recklessroges said one terrorist long time ago...
NJMProductions go plop
Two-minute papers:
"You fellow Scholars somehow started to love the fluids, [....]"
An offbrand-animator researching 'yoghurt physics':
*Sips coffee*
Thank you so much for all the love everyone! 🙏 Our Instagram page is available here with the slowed down post: instagram.com/p/B_5OdwfF_Ur/
The highschool group project law: have any more than 2 people work on a task and it'll turn into a shitshow
FTC Gaming tank crews?
That's why I did all of the work or basically played slave driver to keep the group in line - only to be labeled a bad team player afterwards. Fun times.
Honey fluid was the most realistic one.
@@arsenhere7020 it sayed at the bottom it was slo-mo
@@arsenhere7020 its written in the bottom that its in slow motion
31%
Fluid simulations never look real to me, they always seem like they're happening in slow motion. It may be because there is no frame of reference for how much water i am supposed to be looking at. Like am I looking at 50,000 gallons of water or am i looking at a liter of water.
yoo imagine tsunami simulators now :flushed:
Bright future ahead. Imagine a Tsunami AAA game. We are one step closer to simulating the entire earth in all its glory. Even if not 1 to 1
@@kathoden135 SAO soon will be avaible in real life xd
@@sageofmugen1724 the problem is not the gameplay or world building; that can be done with some hard work. What is a a problem, is the hardware to play It.
@@AndoniOlea xd yes hardware must be run on thousands of Ultra super computer specialized on that game and related with each other through the internet and of course with our VR that can transfert all the information directly into our head through electricity without killing us
YEAH that seems impossible
But if you go back in time to the 1700 and you tell them that will be a talking bright boxes communicate with each others and you can do....)
They will say IMPOSSIBLE
so that's the bright side you can't expect what sciences can do in the futur
haha, flushed
dang I need this paper in blender ASAP
@LapisSea please do
I haven't read the paper, but implementation uses NVIDIA GVDB Voxels, which would be NVIDIA GPU only. In order for something like this to be added to blender it should be compatible with all common GPU's. I'm not sure if you can do sparse grids in like OpenCL. I don't have any experience in GPU programming.
@LapisSea insert linus torvalds fuck nvidia gif
@LapisSea well, I looked around for 2 minutes and the code seems to be open sourece, so no legal problems-yay!
@LapisSea If you actually want to do that, I would like to cooperate on implementing this for Blender. I'm sure it would be way easier with more people involved, and I also have suitable hardware to run tests.
"and runs on your graphics card" well maybe not my graphics card
Buy a RTX 2080 Ti, it will only cost you your soul.
theRPGmaster The 2080 is cheap compared to the Rtx 8000 quadro
@@theRPGmaster Make it 3080 Ti
@@getsideways7257 😂😂😂😂🤣🤣😂🤣😎😎😎😫😫🙄😣😥😴😌😀🥰😅😅
It ran on a Quadro GP100 so you might have a chance I guess..
Please, please keep doing these videos. I love it whenever I see a new Two Minute Papers video in my sub box. I'm by no means a materials engineer or computer scientist, so your excellent, concise explanations of these topics makes it possible for hobbyist-level computer enthusiasts like me to stay up to date with cutting-edge physics simulation. Thank you for your work!
I'm holding onto my papers just reading that title
"Dear fellow scholars"
Me, sitting in my pajamas at 3pm covered in junk food and crumbs: yeah
Is there a contradiction in here? I don't see it :)
Most people is wearing their pajamas at 3pm while eating some junk food this days... 🤷🏻♂️
...
Imagine HL Alyx, but with a shower simulator.
*Valve stock goes through the roof.*
Just wait for the modding tools for your dreams to come true.
Now I'm just imagining Alyx standing under an Aperture Science Cleansing Gel Tube from Portal 2.
Where do I sign up?
I hate to be that guy but isn't Valve a private company?
@@chasegraham246 yes. Why are you asking?
I love your videos! Been watching for quite a while now. I love how short and to the point the videos are, they make it much easier to digest the complex information being discussed. Keep up the great work!
We love your videos a lot. I click it as soon as possible. It is really a treat to know new technologies about simulations and Ai
I've never understood how the graphics card was able to perform different types of tasks faster than a cpu. I finally get it now thanks!
Look at pixel shaders and how they are processed!
A lot of people have trouble writing pixel shader because they think about the big picture too much.
The thing is that your pixel shader will receive one "pixel" coordinate and his job is to say what to do with that pixel.
So, if you want to draw a circle, you should not think about "how can I draw a circle", but, how can I know that the point is in the circle so I can color it.
Then the graphic card will run multiple copies of your shader with different coordinates as input.
Now, imagine you have a list of numbers you want to double. You feed your shader with a picture that is the list, the shader will run in parallel and the picture you get back is actually your list mapped with "x -> 2x".
Where graphic card sucks, it when they have to iterate. The CPU is just much faster to do iterations! But with some trickery, you can transform iterative task into parallel one.
One example is sum of n first integer or Fibonacci series. They both have a formula giving you the n-th ellement without iterating!
Another interesting watch is raytracing shaders. 'The Art of Code" is a good channel for that: ruclips.net/video/k1vlDbhh0ws/видео.html
Think of the graphics card as a massive mapping operation. It can map millions of inputs to millions of outputs easily. That could mean shading pixels, placing vertices, simulating water drops, or whatever.
Then if you want to get *really* advanced, the GPU actually executes these mapping operations in groups, and individual threads within a group can talk to each other and coordinate. There are some really clever algorithms which take advantage of this grouping.
1:37 imagine if there was realstic flowing water in minecraft... that would be so weird.
*Mining*
_Oh fuck here comes_
It would be hardcore x5
Coming Soon.
Need this in GTA 6
That would arguably ruin part of the game since a lot of builds would be ruined since water sources will no longer be a thing
Ikr, terraria already has this, why not minecraft xD
Woooooow! I'm facing that problem exactly right now: I'm working on a simulation of an aircraft splashing down into water, and given the speed of the aircraft and the splash in all directions, the domain is huuuuge. At any vaguely reasonable resolution that means 32 GB of RAM fill up long before the first frame is calculated, the software freezes and no patience in the world will let the simulation finish... so this is gold!!
every time i hear the word real time i think of Dennis Gustafsson who’s making a voxel based destruction game. it’d be really cool to see more games that utilize voxel based physics
Sadly it wont support current realistic 3d graphics techniques as you will now need data for the inside of the object. and that will make game creation mega expensive.
@@ali32bit42 Take a look at UE's Chaos physics engine, lots of potential. And I can't wait until it's production-ready. I'll use it for sure.
@@theRPGmaster i have seen it. problem is fully destructeble environments break level design methods in most video games. and its already very expensive to model the outside of in game props now imagine having to model the inside of millions of props as well. this tech has existed since early 2000s but most devs dont use it because they cant design cool game concepts around it. why would anyone solve a puzzle when they can just break the door ?
@@ali32bit42 have one of the next major npcs say: "I heard you came in here by breaking our door. This will not work. Go find another job or I will call the police."
But yeah I see what your point is. It might really take some further steps for level design to catch up with these techniques. Either with AI that supports the level design or just with new creative game concepts.
@@ali32bit42 Destruction (even if just prefabricated destruction) is always cool in action/shooter games, and really builds immersion. I bet in the next 5 years technology would progress enough so we could use voxel based physics in a whole lot more games since optimization and hardware is progressing super fast. It won't work in all games, but that's alright. We've been making story based puzzle games forever and I don't think that'll change just because of voxel based destruction (or destruction in general)
Soon enough, we'll have processors fast enough to simulate the real world to tackle down different types of global issues. What a time to be alive indeed to witness such amazing physics!
Scientists already use simulations to tackle down global issues (like weather prediction, geological simulations, etc). The only difference is that this method will be a lot faster than current implemented methods for simulations.
I dont think our processing power is that problematic. I think the problem in the real world is information. You can make a scenario on a computer but the more temperature readings we have in the real world the better we can make the simulation if you're looking at weather. Also wind samples and humidity samples would be required too etc.
The problem is we cant see every single particle of air. We have samples spread out over miles between stations. If we could fill the air with sensors, we could get more and more accurate reading because we would have that information of everything going on.
So unless we know all the information about a natural real world system, Our simulation's accuracy will vary. I mean theres probably inefficiencies in building simulation variables but lack of information is the issue, right?
@@DrDinkle it's not necessarily lack of information but more of how the complexity of the problem evolves for computing. If you're serious about finding the answer to your question look into asymptotic analysis (big O, big omega, and big theta). Also look into P vs NP problems. None of this has to do with fluid sim btw.
It's probably both but without the right processors, no one or only few would try to attempt and study the information required. With more powerful processors, comes more demand in powerful ways to use it.
Thanks to ryzen !
20 years ago, long before YT, I used to play with fluid simulators in 3d studio.
for me it was the most gratifying and oddly satisfying experience at the time even at the price of days of compute to get 20-30sec of water waves splashing around.
these days I’m more into robotics and AI but I still watch this channel for news on simulations and I still love it even if it’s not my field anymore.
great work anyway, please keep it up!
My scholarly contribution: Some time back in the early 2000s I did a Cinema 4D plugin that used just the CPU cores (multiprocessing) and it was able to put up to 2,000,000 (simplish) polygonal objects over another object in about real-time, parametrically. As noted in the video: independence of parallel processes is *key*. They did not have GPU core support in their API at the time so I can only envision how much more could have been done.
As I am now diving 'deep' into Machine Learning, I find your videos extremely informative and worthwhile. Many Thanks!!!
Nobody:
3d Porn artists: *It's free real estate*
ohh yees, that's where the money's at!
Oh god
Nobody:
understands the nobody meme
Yeah I'm pretty sure that I've seen this paper put to... Ahem... Practice.
Can’t wait till this comes to blender
Have you heard of the "taichi elements" addon that is currently under development? I would say the "taichi" thing sounds pretty similar to this. Here is a link to a video about Taichi:
ruclips.net/video/wKw8LMF3Djo/видео.html
also the github page for taichi elements addon:
github.com/taichi-dev/taichi_elements
you might as well take your blender license and blend it up in a real blender
@@sharantheboss2957 haha what
That mantaflow bro...
Chemical engineer here, I have worked with grid simulations of fluids... I had always thought nobody liked these subjects, until I saw your channel. Honestly I have to look at that paper but hearing an extended abstract in a comfortable youtube video is priceless
We should be the ones thanking you! Thank you for exposing these game-changer innovations for us in an accessible and fun manner! I'm so excited for the future of fluid sim, and I'm so grateful for people like you, who are contributing to the spread of this information and helping move the tech industry foward!
'You fellow scholars, have started to love the fluids'
hmm...
Holy crap, imagine Splatoon using this.
Splatoon's 'simulation' is already 100x faster than this, and has little need for this level of accuracy. At seconds per frame, this is not a realtime technique.
Games like Splatoon or the goo from Portal 2 opt for a 'good enough' technique that is blazing fast, but only kindof sortof accurate.
This is mainly a technique for rendering accurate, large-scale simulations in days instead of months.
That said, games with this level of detail would be amazing! I believe that pseudo-accurate realtime (
@@sentinelnox6894 I was talking about new game mechanics - paint spilling from pools, containers, or high-viscosity paint slowly flowing like honey. It's slower for sure, but that's what optimization and hardware is for. It would absolutely add a lot to the gameplay.
That cake from 2017 looked amazing, when a simulation makes you really hungry you know it's nailing the realism!
This is a really great paper for two main reasons. The first thing that it does is show the innovation and invention of a simulation that was greatly improved if not created entirely. It also shows how to further the idea by listing its faults and how to get around them. A example is the fact that it requires so much computing power it is confined to a small space. But you can get around that by targeting small areas. All this means is that the paper successfully pushed a subject of science further forward into the future while also allowing other scholars to build upon and improve the simulation and further the cause.
"You Fellow Scholars somehow started to love the fluids, and I am delighted to see that".
*LADIES*
-Hackerman.
last time i was this early i started a fluid simulation with an old method, its still not done yet.
Honestly, you make the content interesting enough that even those who are not scholars, and do not have a good understanding of the mathematics and content, can understand and it makes it so lovely!
You are too kind. Thank you so much! 🙏
Love watching fluid simulations. Thanks for doing the work to get these videos up. It IS appreciated.
2 Minute Papers, AKA simulation porn xD (Something something dirty joke about fluids?)
Hey, I caught you here!
2 Researchers one cup.
Sorry, I'll see myself right out.
@@Megalomaniakaal What combinations of words have I just read with my two holy eyes?
Sam Pulley yo this guy has holy eyes
@@Megalomaniakaal Ah yeah that would be a really interesting 'fluid" simulation.
4:34 "At the risk of oversimplifying this situation..."
Dude. You lost me at coffee in 30 seconds.
What a delight of video. Thank you!
its always a good feeling when people love your fluids.
who doesnt love fluid sims
They're scary to me.
When you say "finally" I imagined it would be a paper that just came out, not from 2018.
I know - haha. And nearly 8 minutes long. FALSE ADVERTISING 😅
Still great, though.
@@jacanchaplais8083 You should ask for a refund, or pay 4 times as much. Same same.
Knowing this paper exists, and slowly learning water sims at the same time is throwing me for a loop because I now know there's a faster way to do what I still don't know how to do!
I'm very thankful for this channel as it allows me to keep up to date on technological progress.
If I have the choice, I will choose to spend all my 9 lives working on science & technology 🖖
Who wouldn't?
@Toufiq Imrose anti vaxxers are speedrunning their 9 lives
“Dear fellow scholars”
Me a high schooler: 😐
Soon 😁
A high scholar*
Me, an 8th grader:
: |
Katskull I’m 8 too
@@decodedbunny101 8th grade is not being 8, buddy.
Thank you for providing such an invaluable public service with these videos!
I know absolutely nothing about programming or computer science and have no interest in learning, but I still watch every one of your videos.
but can it run on my computer
That's the whole idea of this video unless your pc is a potato ,like me.
Depends on the gpu you have.
can't wait to see the impact on biology research
None. You could already achieve these result (it was just 20x longer) the real impact will be on video games
@@auxencefromont1989 to say it will have no impact is a little harsh, being able to simulate fluid faster means you can change variables and re-render more so while i dont know much about biology personally, being able to run more varied simulations has to be beneficial, also anything that improves gaming improves training for example as vr improves we might get to the point where surgeons are able to simulate a surgery and run through it 100s of times in a day before performing that surgery, im sure you can see how that could be beneficial
@@auxencefromont1989 Video games already use the graphics card.. Nothing new here to help video games.
I absolutely love these videos! You're clearly genuinely excited by the progress made in the fluid simulation field, and it's a subject I'm really interested in too. Your explanation of the drawbacks of computational parallelism was intuitive and accessible. Keep doing what you're doing! You've got 1m in your channel's future
I'm surprised people don't like the fluid simulation videos. They're some of the most interesting stuff you put out on this channel, IMO.
Do you say a new unique instance of “what a time to be alive” or do you use the same recording?
Theodore Yamada-Dessert maybe he uses a stream deck. Also “hold on to your papers...” could be on it :D
Well, ofcourse we like fluids. After all we get all of our oxygen from a fluid
I am a graduate student in geology. Fluid simulation applied to geosciences would be an amazing tool to understand the dynamics of rivers, coastal environments, landslides, volcanoes and so much more! I can't wait to see it, and to try using these tools myself. With these faster algorithms we will soon be able to work with large scale modeling, of entire depositional environments. Amazing.
Literally thought you were trolling us by pouring honey onto a white floor. Completely lost my sh*t when I realized it was computer simulated. Amazing how far technology has come since the age of flipping switches to do simple addition.
Scholars: Let's run graphics on a graphics card!
Me: ....
I love your videos man, I hope you continue to do more. This stuff is mindblowing to me
Superb work, thank you!
One of the guys who worked on this published a thesis paper a year after this on mesh based physically rendered knitting. It's absolutely incredible, down to the smallest fiber without needing absurd rendering farms.
Thank you for making videos! I'm a film student and know nothing about computer science, but I love watching them and seeing what the frontier of simulations is like!
You rock! I'm not in this field. Your descriptions are brilliant, making the subject quite accessible for me. Thanks!!
3:07 a nice example of three different types of unicorn excrement
Thank you for these wonderful videos!
I wish I could love this brutally authentic and vulnerable video.
I didn't understand that much of your explanation but waawww I didn't know it was even possible. These simulations are incredible and fascinating!! Thanks for sharing
Incredible video! I can't pretend I understand most of what was discussed, but it's incredibly fascinating!
These fluid simulations give my cervical vertebrae a weird tingly feeling. It's so damn satisfying.
I personally don't care much for fluid simulations but I have really enjoyed watching the progress in the field over the past few years.
Nobody showed up to _fluids simulations_ videos? But that's one of the most fascinating (and satisfying to watch) kinds of physics simulation!
I've been enjoying watching your older videos from time to time, now that you reminded me about them I'm going to do it more :) Thanks for keeping this up through the years, you deserve the popularity!
im not exactly sure why, but that happy bounce at 4:16 was somehow very satisfying to watch!
Appreciate the effort and time taken man, thank you!
Love to see all the different kinds of things you post. For me the excitement comes from how it inspires my imagination for what I may be able to see done in real time just a few years down the road in the many video games I play. Keep up the great work!
Fluids are getting so much love the past few years and a lot of the papers feature open source software too.
Great stuff.
Hey Dr., the videos on soft-body and fluid simulations are some of my favorite. Thank you as always for the awesome video!
Subscribed. I enjoyed watching your previous fluid simulation videos, but haven’t subscribed till now for some reason. Liked and shared. it’s kinda wild we’re all living in the future right now
Stunning work! I enjoy these 2 min papers a lot! Great videos keep it up! What a time to be alive :)
Awesome, Production shots are now possible, not only liquids filling containers!
i am actually really really new to programing but i love these videos and you can learn a LOT from them
This is such an awesome video concept because I don't have to be an expert to understand most of the key ideas! I wish there were more researchers from other fields who present their awesome work in such an extremely simplified way.
This must be the other heartwarming message of the day. Thank you so much! 🙏
Dude, this is why my free time never gets boring. Keep up the good work, very interesting.
yeah, things thats been used for science and research finally come to pc gaming
performance still mystery though
These kind of videos and thought experiments are important. Sharing them helps future scholars or even something as simple as curious children push the technology further in their future. Math, technology, and a push for progress do not happen in lonely, isolated environments. Thank you for sharing. Knowledge should never be lonely.
1:41 - YES! Minecraft totally needs this!
Your videos are awesome, I love to see the fluid simulations you make and to hear your comments about it!
After the information age, comes the simulation age. So much information, so many simulations, so little time.
Fascinating how our understanding of math and laws of physics can reproduce a realistic result!
6:38
man that honey drop is literally more satisfying than ''satisfying videos''
You have no idea how happy I am that I found your channel. 100/100 videos you upload are interesting to me. 😂
I really Appreciate your Handwork
Wow! This is incredible!!
thank you for posting these, and thanks to the algorithm from showing them to me. i'm not really hardcore into these simulations as you are, but the high order of calculations fascinate me.
2:44 The most appetizing thing I have ever seen! 😮
I have always found fluid simulations fascinating, since my early days watching old NVIDIA demos or the PS3 tech demos.
WOW ez nagyon jól néz ki, imádom a videóid, nagyon informatívak!
Thank you so much for what you do for the community.
Thank you so much for your kind support Gammett!