How hard is it to optimize the current "engine" of VaM 1.X

[haroldwithlove]

Sorry for this layman question. For a long time, I thought VaM stopped updating until VaM 1.21 was just released (which surprised me). So it seems that improving the performance of VaM 1.X is still ongoing in some way.

I have very poor knowledge of game developing, but I wonder how hard it is to further optimize the current VaM 1.X so that it utilizes more cores of CPU and GPU, like what is promised in the development of VaM 2.X. I know this is not an easy job but could any professionals explain how hard it is to do such optimizations based on the current VaM 1.X? It could be a project parallel to the development of of VaM 2.X...

 

[SlimerJSpud]

The simple answer is they don't have enough human bandwidth to do parallel development of major enhancements on 1.x. Meshed decided to make a 1.x update probably due to popular demand that they fix some of what's called "low hanging fruit." Back porting a major part of the 2.0 enhancements into the 1.x "branch" would be a lot more work than they have bandwidth for. There was another 1.x update a little while back that improved loading times somewhat.

 

[MacGruber]

but could any professionals explain how hard it is to do such optimizations based on the current VaM 1.X? It could be a project parallel to the development of of VaM 2.X...

Meaningful optimizations would need to do two things:

1. Upgrade to a more recent version of the Unity engine. VaM 1.X is using Unity 2018.1.9, which released in summer 2018....in engine-tech terms that might as well be a century old.
2. Better utilization of available CPU cores. VaM does a lot of things in the main thread of the Unity engine.

Both of those things require an immense amount of work, large parts of VaM will have to be rewritten from scratch. Meshed calls that project VaM 2.X

 

[DJ]

It's worth adding that as well as being an immense amount of work, moving beyond Unity 2018 would completely break every existing Vam scene.

 

[qoosh]

This is why 1.X will always be 'last-gen' because it is based on a 'last-gen' engine. But I like what creators like PluginIdea are doing with the volumetric lighting, and other performance enhancing plugins and features that they are putting out. Hopefully they can continue to work on optimizing 1.x as much as possible, since 2.x is a long way away.

 

[hazmhox]

I have very poor knowledge of game developing, but I wonder how hard it is to further optimize the current VaM 1.X so that it utilizes more cores of CPU and GPU, like what is promised in the development of VaM 2.X. I know this is not an easy job but could any professionals explain how hard it is to do such optimizations based on the current VaM 1.X? It could be a project parallel to the development of of VaM 2.X...

Rule of thumb: if you start a sentence with I have very poor knowledge of something, avoid imagining in the same sentence what could be done.
Wait for the answers first, then after you have all the keys to understand, you can imagine stuffs based on that

That's like the average gamer on steam forums, "I have no idea what gamedev is but they most definitely can fix this tomorrow" x']

@MacGruber summarized this extremely well. It's hard, and not doable if meshed wants VaM 2 to exists in a close future.

Another note, this has been mentionned several times on the discord: no matter how hard you will work to optimize VaM (1 or 2), the goal is still to get an extremely accurate physical simulation. Never picture VaM becoming a game running at 120fps, with dozens of characters in a scene, animated, with soft bodies, realtime simmed hairs and all that jazz.

VaM will stay in the realm of 2 to 4ish characters, the rest will be non-sim characters/enviro if you want to keep proper performances. And since VaM 2 is targeting a really higher standard in visual fidelity and simulation... the experience will (might) be better with two characters, but if you use the highest resolution characters, put more than three of them, with crazy hairs, clothes, sim everywhere, settings at max... you will still likely be in a similar framerate situation as in VaM 1.x

( Just saying that, so that everybody has proper expectations when it comes to VaM 2 )