A letter to the CEO of Unity, Mr. Bromberg,

Unity Engine
, , ,
1

Dear Mr. Bromberg,

I hope this message finds you well. I am writing as a concerned member of the Unity community, wanting to share candid and constructive feedback in light of recent developments around Unity’s roadmap.

Unity is currently navigating a challenging phase—particularly following the “Runtime Fee recovery” issue, coupled with the underwhelming reception of Unity 6. As it stands, Unity 6 is a rebranded 2023 LTS with Resident GPU drawing added in. While this feature is a step forward, it does not represent the kind of transformative leap that could restore Unity’s edge in today’s competitive market. Moreover, looking ahead, Unity 6.1 does not appear to offer substantial new capabilities that would reassure developers of Unity’s long-term viability.

With large-scale projects often spanning years, the absence of features like frame generation and realtime path tracing makes it difficult to argue in favor of Unity as an engine of choice. I therefore urge you to establish a dedicated task force, aiming to implement these technologies within the next six months. One might see this unnecessary, as it is not difficult to see the status and value of such technologies in today’s market. However, as I stated above, projects take a considerable time for its fruition, even years. A year, or two down the line, I can confidently project the weight of the technology, regardless of my personal liking. Unfortunately, while end of a project is easily a few years out, the beginning of a project is today, and that is a problem that we have to consider. Most Unity projects released today began with what Unity 2021 or 2022 had.

On that note, I was disheartened to learn via the forums that Unity’s frame generation efforts—reportedly close to completion—have been paused. I believe this is a significant mistake. Frame generation stands as one of forward-looking features, and leaving it in limbo risks harming both Unity’s reputation and the confidence of developers who need modern rendering options to keep their projects competitive.

Regarding the Unified Renderer, I understand the desire to finalize it before allocating significant resources to new features. Yet, while the Unified Renderer is undoubtedly important—much like perfecting a store’s entrance—it should not overshadow the urgency of modern rendering capabilities. Addressing DLSS4, Frame Generation, and real-time path tracing is equally and if not more important. I say this because Unified Renderer is a efficiency problem at its core, while DLSS4, Frame Generation, and real-time path tracing, is a door Unity users cannot enter at the moment. Yes, the industry might not move in that direction, it might not be fast enough or good enough, but that choice should ultimately be the user’s choice, not by Unity.

Today, DLSS4—or, more precisely, multi-frame generation—was announced, and from all indications, Unreal Engine will support it from day one. That means Unity is now at risk of lagging two generations (RTX 40 and 50 generations) behind by the time any integration occurs. Even as you continue refining the Unified Renderer, it is critical that these advanced features not remain on the back burner. The community needs to see clear progress on both fronts to ensure Unity remains a compelling choice for studios and developers over the long haul.

I write this out of genuine concern and in the hope that Unity can regain the community’s trust by delivering real solutions in a timely manner. Many of us still believe in Unity’s potential and want to see it thrive. I appreciate you taking the time to hear me out if somehow this post finds its way up to you, and I hope you will consider these suggestions as you guide Unity into the future.

Sincerely,

-Peter

2 Likes
2

I have some thoughts on this…

You mention several times frame generation and real-time path tracing. So based on the number of times you talk about it, I have no doubt that this must be pretty high up on the priority list from your perspective.

But that’s just the issue… the items you mention are high priority for you, but is it high priority for the majority of Unity users? I don’t know… I have no data on this obviously.

I also have a list… SSGI that doesn’t break on foliage and moving character skins due to noise and other artifacts, that has more toggles and options to control its overall intensity and GI / occlusion contributions. I would love a built-in HDRP terrain setup that let’s you choose how many layers you can draw in a single pass, instead of being hard-coded to 8. I would love to use Shader Graph to create terrain shaders. I would love for long-standing bugs to be fixed in LTS versions, breaking stuff that has been sitting there for months. SSR with more advanced options would be nice. Built-in support for a wider range of upscalers instead of just DLSS and FSR1.

I know at least of some people who would love to see a wide range of issues fixed with XR+HDRP.

There are so many core priorities to deal with, that it makes real-time path tracing seem like a distant niche thing where (probably) just a minority of people would like to see this as a higher priority than the core stuff.

I think their focus on Unifying the render pipelines is a necessary and urgent step, as the split between HDRP and URP has caused untold harm by splitting their own efforts regarding graphics pipeline development, as well as splitting the community and asset store in two. Laying these strong foundations will be a good move.

3 Likes
3

George,

While I agree with most of comments, and feel the same way, there are some key elements that I would like to point out.

  • Most of what you mentioned: SSGI, HDRP terrain, Shader Graph integration and advanced SSR, can be somewhat worked out via asset store (SSGi → Try H Trace, HDRP terrain → Try InTerra, SSR → I don’t have one XD) or even can be worked out with your own custom solution at the moment, it may not be as good as you’d like, and it could be time consuming but it can be somewhat remedied.
  • The Frame Generation and Realtime Pathtracing (with Ray Reconstruction) requires Unity to actually work on it as asset store cannot cover this, (Frame insertion within the current Unity is somewhat not possible) and Ray Reconstruction + realtime pathtracing (I have no idea where to even begin). AFAIK, there have been attempts to implement this by asset store devs, and me personally, but ultimately get stuck.
  • Unified Renderer, in my opinion is something that needs to be done, but at the same time, it shouldn’t pause Frame Generation and RP tracing integration. Because, let’s say a good 2 years until Unified Renderer becomes production ready, another year for FG + Realtime Path tracing implementation, we are talking about 3 years. Then devs will plan projects using this tech which let’s say on average take 2 years until release, we are roughly talking about 4~5 years from today until we see a game with frame generation and realtime pathtracing using the Unity Engine.
  • Frame generation and realtime pathtracing is hardly a niche area. Yes, it might feel that same way as gamers today, but not as developers planning a project today. We always have to work with future in mind as projects take time. Whether we like it or not, this is where industry is headed. Most of rendering replaced with accelerated AI equivalents.

I am not trying to say that you are wrong, but I just felt like that with the latest pause on Frame Generation efforts within the Unity dev’s and the new 50 series announcement, I felt that this could be more damaging than one might simply think.

The worst part, is that most of the foundational work is done already, DLSS3 is done, most of Frame Generation is done (I heard from the dev that it is near completion) and there is already path tracing available. The executives literally have to form a small team to tackle this matter, simply restructuring the HDRP pipeline to bind nVidia APIs. Literally, this is a low hanging fruit that completely alienates AA and AAA level usage of the engine.

The quicker Unity can set a date or plan of the feature
The less the damage, and more flexibility of devs.

Or so I think.
Hence the letter to the CEO
I hope this helps.

1 Like
4

Do you have his email or something? Doubt he’s a frequent guest here.