Search Unity

At Unite Copenhagen, we talked about what’s on our roadmap for our upcoming Unity 2019.3 and 2020.1 and presented our vision for what you should expect in 2020. Check out the video of the talk below, and read the finer details in this PDF copy of the slide deck from the talk.

Unity’s evolution is guided by four main themes: core and performance, workflow and creation, quality and fidelity, and deploy and operate.

Core and performance

With the Data-Oriented Technology Stack we are rebuilding the core foundation of Unity enabling you to get optimal performance on a variety of hardware. However, DOTS performance not only unlocks what you can build but also how fast you can build it decreasing iteration times significantly so you can be more productive.

Workflow and creation

We’re laser-focused on improving workflows. That means better integration of artist tooling and better productivity for teams.

For example, with our new conversion flow for Data-Oriented Tech Stack (DOTS), you’re able to use the same intuitive content authoring tools that you’re used to with GameObjects, Component and power tools such as Nested Prefabs, while at the same time taking advantage of the massive performance improvements that come with runtime optimized DOTS data. At the same time, Unity Live Link will improve your iteration time drastically, with direct feedback on the actual targeted device.

Quality and fidelity

Bringing the Universal Render Pipeline to parity with our built-in renderer and revolutionizing performance on a variety of target platforms is a huge part of our focus. We’re also concentrating on getting the High Definition Render Pipeline (HDRP) production-ready and building in real-time ray tracing. Our real-time raytracing solution is a full-frame path tracer with ground truth Global Illumination, built on top of HDRP for the best possible quality in a real-time engine.

This means that with Unity, your project will look its best, and perform optimally – wherever you find an audience.

Deploy and operate

As you find an audience for your Unity projects, we will be there at every step of the way to support you. We offer the broadest platform reach, global content distribution and services to create increasing opportunities for player acquisition, engagement, and retention.

We’re thrilled to share that platform support for Stadia will soon be production-ready for approved developers, beginning with Unity 2019.3. This build will contain support for everything you need to ship your first game on Stadia, inclusive of Unity features like the Data-Oriented Technology Stack (DOTS) and High Definition Render Pipeline (HDRP), plus exciting features unique to Stadia like Stream Connect, State Share & more.

Stadia Early Access Program

Our Stadia team has worked closely with and gathered feedback from developers in our Stadia Early Access Program over the last year to craft a familiar deployment experience. We’ll continue refining this experience based on user feedback.

You can reach out to Google for more details on how to apply for Stadia developer access by visiting the Stadia Developer website.

In the meantime, we have a few suggested steps your team can take now, in parallel to public versions of Unity, to have better “Stadia readiness” (should Google provide you access to dev resources).

  1. Pull the 2019.3 beta for a preview version
  2. Build for Linux (3rd party dependencies without Linux support should show up here)
  3. Use Vulkan only
  4. Start using IL2CPP (available for Linux as of 2019.3)

As always, sound off in the comments and let us know what you’re excited about, ask questions about what we have planned, and what you’d love to hear more about. Thanks to everyone who attended the talk and to all of you for your support and feedback – it’s what helps us make Unity the best game engine it can be.

Note: we are currently working on redeveloping our public roadmap website. In the meantime, please consider this presentation our most accurate representation of our plans.

23 replies on “Unity Roadmap @ Unite Copenhagen 2019”

DOTS isn’t a magical solution. The vast majority of my CPU time ISN’T on gameplay code.

Level of Detail? Horrible and monothreaded.
Camera? Massive overhead for each camera.
Physics? Overhead per collider regardless if they never move

Instead of telling people they should rewrite the gameplay code, perhaps Unity should start by having their own code more optimized.

And that’s why their rewriting basically all the base systems in DOTS. Even if you don’t implement your logic in ECS you should still benefit of, for example, having renderer culling and mesh skinning be multithreaded and cache-coherent

That is exacly what they are doing with the already-available multithreaded rendering (look at Vegetation Studio Pro asset if you want a good example of managing level of detail through burst compiled jobs), the DOTS Physics preview package, texture streaming & virtual texturing, as well as all the work they’ve made in their scriptable rendering pipelines. Everything you just mentioned is a reflection of how little you’ve read from their roadmap.

Great update! Thank you for keeping us in the loop.
Unity Live link for code – uh, I feel all sorts of jitters in my stomach :) Yes, please!

Looking forward to the new improvements! But the current GameObject/MonoBehaviour system and the current GUI system aren’t being dropped any time soon, are they? I’m happy to use DOTS for my next game, but I can’t see myself transitioning my current project to DOTS (especially because it’s already built upon another ECS library).

Very much not. Not any time soon. Especially considering the current editor workflow for DOTS is using MonoBehaviours as the editor-side data containers.

This PDF + presentation video + blog post = an awesome thing!
Most people ARENT at Unite, so it’s great to get this information in a timely and concise manner. (Especially the detailed PDF.)

ALSO: “we are currently working on redeveloping our public roadmap website. ” — Thank you!!

Any updates on when 2D sprite support will be added to DOTS? I’ve seen forum posts about homemade solutions using quad meshes to render sprites, but I was wondering when auto-conversion for the SpriteRenderer component will be added.

It was mentioned in roadmap briefly. It is DOTS only and if I recall correctly it will land as a preview in 2020.1
You can download an experimental version on the forum.

I quite like the new format of the Roadmap talk. But I would prefer it if it came at the same time as a new web-based roadmap that properly covers the modern modular Unity era.

Good but no mention of webgl ever? Webgl seems a very risky platform to develop for with Unity yet we get no news about it.
On mobile it is, at least by default, unusable.
On Windows it is ok if you are careful with memory.
On Mac, Safari likes to popup messages about memory and ‘energy’ use and suggests you close the page so not a good look.
I know you are limited by web support and Apple’s in particular but updating people about the future of webgl with Unity and suggestions on where it is still usable would help.

Hi Andy thanks for your question. We actually answered a question on this at the end, but to take a step back – we were not able to cover a lot of things in this talk as we simply feel that the format we had previously where we would list endless features isn’t of benefit to helping you understand why we are doing what we are doing – I really hope this felt more thematic and helped you understand our approach.

Back to your question – someone asked about WebGL during the questions at the end of the talk as you will have seen (I know you of course watched the whole thing, it’s like a book you can’t put down ;) ) but to reiterate our position on WebGL – we are at the behest of standards developing which of course whilst we partner with many of those involved in driving the standards, we also cannot build what is not supported. We are pushing heavily on our Project Tiny part of Unity – effectively a DOTS runtime driven way of taking your game to HTML5 in 2D and soon in 3D as well. This is currently our most size optimal approach to delivering web based content.

This, combined with our ongoing support for WebGL itself means we are continuing to invest in this area, so although we did not call out WebGL in detail in the talk, it be no means.. means that we are not working on it. Thanks for your question – sorry we cannot share more specifics at present, any follow ups, let us know.

Thanks for the info Will, and just a further question – do you know if Unity will be expanding ARFoundation to WebXR when the standards are defined later this year or next year? Thanks!

Comments are closed.