Vulkan Support – Future-Proofing Second Life for Better Performance & Graphics
tracked
KiraYoichi Resident
🔴 Summary 🔴
Second Life has come a long way, but OpenGL is becoming outdated. To ensure SL remains visually competitive and runs smoothly on modern hardware, I propose that Linden Lab begins development on Vulkan support as a long-term goal. This transition would greatly improve performance, reduce crashes, and allow SL to take full advantage of modern GPUs.
🟡 Why Vulkan? 🟡
✅ Better FPS & Performance – Vulkan is optimized for multi-core CPUs and modern GPUs, meaning higher frame rates and less lag in complex environments.
✅ More Stability & Fewer Crashes – Vulkan manages memory more efficiently than OpenGL, reducing viewer crashes and graphical glitches.
✅ Future-Proofing Second Life – OpenGL’s development has slowed, while Vulkan is the industry standard for new and upcoming graphics engines.
✅ Improved Graphics Potential – Vulkan supports advanced rendering features that could enhance lighting, shadows, reflections, and materials in SL.
🟢 How This Transition Could Work Smoothly 🟢
Instead of a sudden shift, I suggest a gradual development plan (2025-2030):
1️⃣ 2025-2026: Linden Lab researches Vulkan feasibility and starts experimental development.
2️⃣ 2027-2028: An optional Vulkan beta mode is introduced for testing and optimization, running alongside OpenGL.
3️⃣ 2029-2030: Vulkan becomes the default renderer, with OpenGL as a fallback for older systems.
4️⃣ Community Engagement: Regular updates from Linden Lab on progress, plus support for third-party viewers adapting to Vulkan.
🔵 Why Start Now? 🔵
Even though this transition will take years, starting early ensures SL stays ahead rather than falling behind other virtual worlds. A well-planned Vulkan integration could attract new users while making SL smoother for current residents.
If you agree, please upvote and share your thoughts in the comments! Let’s show Linden Lab that the community is ready for a modern and optimized Second Life! 👍 💬 �
Log In
Hawk1992 Resident
I wrote a similar post. We need better lighting. Make it be heard! before implemetning any other Mapping related features.
ST33LDI9ITAL Resident
Mmm.. or a renderer that is platform agnostic. DX for Win, Metal for Mac, Vulk/GL for Linux. Tbh rather have DX over Vulkan. Vulkan isn't gonna be the end all solution to problems SL faces and may even complicate things further. It's not a simple port, it's not simple to multithread, and you're not going to see crazy improvements to fps given the way SL is structured just by switching to Vulkan. At least with DX there are suitable translations for Mac and Linux (GPTK and Proton). So then you'd have first class support on Windows at the least which is the majority of userbase and realistically GPTK and Proton have come quite a ways and are plenty sufficient for most games especially if you put in the compatibility work. It would be far less work to port to DX and rely on GPTK/Proton for Mac/Linux.
Hadet Sonnenkern
I think one of the major holdups for this will likely be MacOS users. Though they could do what a lot of companies do and just package moltenvk with the release.
Orion Greymoon
This is an excellent idea. The timeline is concerning, however.
There is popular simulator software that transitioned from OGL to Vulkan in about 18 months. With a pretty small dev team. And the graphical demands and integrations within the software is significantly more extensive than with the SL viewer.
KiraYoichi Resident
Orion Greymoon I gave the time line as a suggestion really, if Linden Lab is prepared and capable of delivering Valkan less than 5 years. Wonderful 👏
Also, time for residents with older computers to prepare themselves.
primerib1 Resident
Orion Greymoon Do note that Vulkan is MUCH lower level than OpenGL.
A 20-line OpenGL code can result in HUNDREDS of lines of Vulkan code.
Someone mentioned things one doesn't concern oneself with if using OpenGL https://www.reddit.com/r/vulkan/comments/115sm77/comment/j9bj3mg/
Quote: Things like swapchains, memory allocation, device management, synchronization, queues, resource pooling, descriptors, etc.
Sure, if someone is
dedicated
on porting things from OpenGL to Vulkan, 18 months might be enough. But I think behind the scenes the simulator software's developer have their own middleware/engine on top of OpenGL, thus enabling easier 'swapping' of the lower-level renderer from OpenGL to Vulkan, while still allowing the rest of the team to develop the same way.And that is the question: How is the renderer of SLV implemented?
Orion Greymoon
primerib1 Resident I'm not a low level graphics dev, so I don't speak from experience. However, I have friends who are. And, as an active user of the flight sim software, very aware of timelines and such.
Are the APIs and code needed to drive Vulkan more details? Yep, I've heard that also. However, that's one reason that Vulkan is more flexible...ability to manage / control the environment in a more granular and performance manner. A loose analogy may be interpreted languages versus compiled. Or VB versus C (and we all love C, right :) )
If there is a layer between OGL and the hardware currently, that does complicate the recipe a bit. Yet that additional layer probably contributes to lower performance / lags / maintenance PITA, etc. Refactoring is never a pleasant...especially since our future looking crystal balls tend to be a bit...cloudy.
primerib1 Resident
Orion Greymoon Rather than directly accessing Vulkan, why not rebuild the viewer on top of Vulkan-compatible 3D game engines that can handle the rendering?
The mobile viewer is already being built on Unity; whatever learnt from this exercise can be applied to a PC-porting onto Unity.
Or alternatively if one does not want to be entrapped by Unity when they again pull out their licensing shenanigans, can try to build upon Godot 4 instead, since Godot 4 is MIT-Licensed.
Both Godot 4 and Unity can drive Vulkan.
marxman1313 Resident
Already been stated but this is the long term goal with the most recent string of updates. We're a ways off still but getting there. I totally agree though.
Vincent Nacon
This is already part of their long-term goal. GLTF and PBR are basically their stepping stones toward Vulkan. There are some few others they need to update before making the big switch.
Mind you, this is no small feat. They will get there, but that will take a while. We're probably looking at maybe 2-5 years, if all goes well.
SL Feedback
Hello, and thank you for your detailed and passionate feature request regarding Vulkan support for Second Life. Another resident has previously brought up the idea of integrating Vulkan API, and we are merging your comments to get to it faster. We recognize the potential benefits you’ve outlined, such as improved performance, stability, and future-proofing Second Life. While we have no estimate on when this might be implemented, please keep an eye on future updates. We appreciate your input and hope you continue to share your ideas to help improve Second Life. Thank you!
SL Feedback
tracked