Mindblowing water...

When we will have something like that?

Not for a very long time
That example is pre-simulated and looped, it uses a massive amount of disk space and memory to render. Since the use case is so limited (in this case it works for a shore line since the animation loops easily) there’s no reason to pursue that kind of feature. Realistically we will have to wait until physics systems are powerful enough to do it in real-time, which is being worked on by companies like Nvidia.

It’s possible to fake that within unreal engine today. A tech artist should be able to put something together using material trickery, but would be a lot of work to fake caustics, find a good way to create waves, faking the layered foam with flowmaps, etc.

Waves and underwater caustics are relatively easy, especially with tessellation and light functions. That crashing foam by the shore is almost impossible in UE4, though.

Looks cool, but video description seems slightly misleading. It is not a real-time simulation, but a real-time rendering of a simulation. Nothing revolutionary here. As you can see, mem budgets to store only 5 second loop are insane. Dubious that anyone would take that into production. With current technology development speed, our grand children will be able to enjoy this kind of real-time simulations in games, after their retirement. lol.

It won’t be too long before those kinds of sims are common place in real-time. Moore’s law is still holding up pretty well…

We’re approaching the theoretical limit for computing with electricity. Below 3nm, electrons will jump bridge and short-circuit microtransistors. We already have 10nm CPUs in production. GPUs are currently being fabricated at 14nm right now. Quantum computing would need to become a thing if we’re going to break this limit.

I would like to see some form of wave crashing particle simulation in UE4 sometime in my life. It would be smart to keep water in the ocean as a heightfield using gerster waves or texture-based displacement that curls in XYZ. But somehow, this should break up into fluid particles at the shore to form the dynamic crashing foam and splash effects, that we also hope will influence gameplay. Making something like this work on a small scale is very different than an open world game, where certain techniques can no longer be used. It’s no easy feat.

Don’t worry, even when they hit the quantum tunneling limit, they will just start building 3d (which they have already been experimenting with for years now). Oh and quantum computing, at consumer level, is still probably several decades away because they would need some room temperature level solution.

This video is made with Unity, that should be possible to do it with UE4 too. I guess.

Probably for a short shore line like in the video. With a very dense tesselated mesh and a detailed map to drive it, plus some particles for the foam…
But then again: who could afford a million polygons per meter of coastline. And this scene is without post processing effects, etc.
you dont have any reflection of objects other than the sun and sky. So its fairly optimized to afford that water.

There are a lot of amazing demos that come from Unity, but those demos are all small-scale. It’s cool to have fully dynamic caustics for a small swimming pool, but scaling that up to an entire ocean is another animal altogether. The techniques used (vertex displacement, render targets) may not work on a larger scale that would make it more suitable for use in gaming. This is why we haven’t seen large-scale fluid simulations in games. It’s all smoke and mirrors.

Indeed, there is a repetition on shoreline and foam.

But, we have our Epic’s wizard Mr Ryan Brucks and his approach that is neat: https://youtu.be/OQ3D0Q5BlOs

:wink:

Not quite sure what the issue is, this is a pre-baked simulation, probably in Houdini. You can easily export Houdini simulations as textures and play them back via a material. Take a look here: https://youtu.be/zGmElCpxZnk?t=1039

ahhhh smart thinking… this would explain why the coastline is fixed and cannot dynamically fit any shame.

That’s really cool, but what is the status of this project? The video is one year old.

Yeah, we know it’s a pre-baked simulation, we were talking about doing real-time simulations though…

I wish I knew. You can ask him in twitter as @ShaderBits and also check his other stuff posted there.