Martin Fuller

Hear how the collaboration between @idSoftware and Xbox ATG brought DOOM: The Dark Ages to life 💫 Learn how they used Variable Rate Compute Shaders to reduce unnecessary pixel work in compute heavy rendering passes, locking 60 FPS across consoles and capable PCs: developer.microsoft.com/games/articles…

@SebAaltonen Consoles and yes, no second 5090!

"The big win in the future, besides quality, for Neural Rendering and DLSS 5 would be in saving production time. But for that to happen, we would need to support all users, and very few people have access to the hardware required for DLSS 5." Agreed 100%. All users = consoles.

Regarding the DLSS5 controversy. NVIDIA insist there is no change to assets. However if all the "improved lighting" the AI has been trained on happens to feature super models, then you are going to get super models as a by-product of your "improved lighting" training.

@JamesStanard I feel the same way tbh, 30 is too slow now even for slower paced games, 40 is the baseline

@MartinJIFuller I’m a snob now and won’t play anything at 30 FPS, but I can definitely tolerate 40 FPS for slower moving games like RPGs and stealth action. And if it enables ray tracing, then that’s a big thumbs up!

Great to see more 40hz support on console. For this type of game I find 40hz to be the sweet point, better visuals than 60hz, more responsive than 30hz.

@NOTimothyLottes AS building is a good thing for async compute to do while rasterising the primary view. Another good argument for raster+RT being the way forward. Plus AS building will be getting cheaper in future

@MartinJIFuller My mind translates this to "low nicotine cigarettes lets you smoke a lot more" ... sorry still not interested. Also RT has nothing going for it in the dynamic world direction, true massively dynamic stuff is too expensive due to acceleration structure rebuild costs.

Haha, "10x RT perf", remember kids, if your frame takes 16 ms, and 4 ms of that is RT, even if RT is 1000000x faster, it's still a 60 Hz game. Rather have them make the rest of the GPU faster.

@NOTimothyLottes @theredpix @idSoftware Personally I’d take the larger play area, fully dynamic lighting including GI, increased geometry+texturing fidelity, physics objects & etc.. of IDTech8

@theredpix @idSoftware I admire the idSoftware folks, but apples to oranges, the total frame time of 2016&Eternal was way lower than idTech8 [won't find idTech8 running on Switch 1 for example]. Would be nice if someone would pickup the 240 Hz baton from where idTech7 left off.

#rendering @idSoftware moved to #Deferred rendering for Id Tech 8 as it proved more performant than their #Forward+ solution. About 26% faster opaque pass #gdc #gamesev

@NOTimothyLottes Generally speaking ;)

@MartinJIFuller Actually I don’t have to factor in any of that raster complexity because I also don’t use triangle raster either.

Lets talk of the universal truth of "Ray Tracing" - RT HW traversal converges to worst case memory access patterns as scene size increases. End of Story.

@SoftEngineer @idSoftware @DOOM The sample is under Xbox NDA unfortunately. It takes advantage of some additional Xbox specific low level optimisations not mentioned in the talk. If I find time I’ll port it to PC perhaps.

About "But we do lower the frequency of shading certainly where you don’t notice." - I did think that the foveated bit was great. Definitelly going to steal that idea if I ever get around to VRCS for my own engine. Do you happen to have a link to the engineering sample that was mentioned in the presentation per-chance?

Watched a brilliant presentation from @idSoftware guys at GPC on @DOOM "The Dark Ages". 11 out of 12 pixels are fake. Say your screen is 4k (3840 x 2160), here's what the game does: DLSS/FSR in "Balanced" mode upscales from: 2227 x 1253 -> 3840 x 2160 (your full 4k) Before DLSS gets the image, it is upscaled by DRS (Dynamic Resolution Scaling) internally from, say 75% resolution: 1670 x 939 -> 2227 x 1253 Before that, the image is rendered with VRCS (Variable Rate Shading) with ~43% of pixels actually rendered and the rest just filled in. So in total we go from 674k pixels actually rendered by the engine, to 1,568k after VRS to 2,790k after DRS to 8,294k after DLSS For a grand total of about 1 in 12 pixels actually being rendered, the rest? - they are fake. source: youtube.com/watch?v=mvCoqC… as for me - I'm not against the fake pixels, bring them on I say, as long as it's done well.

@SoftEngineer @idSoftware @DOOM Thank you Alex. My take is simply not to be wasteful of GPU resource. Spend it where you see a benefit. Not where you don’t. The fundamental truth is not all pixels are equally important.

Hey Martin, loved it actually. I did find your dislike of noise somewhat amusing😅. Yeah, I did like the part that you guys did depth (viz) at full resolution, and then feed it as one of the inputs to compute RSI. In terms of quality I think that's a really good call. I find it a bit surprising how many people rail against "fake pixels", and @DOOM The Dark Ages received high praise for "clean" image at high FPS even on lower-end devices. After watching your presentation - I thought I'd chip in on behalf of fake pixels. As we move to larger screens and higher DPI, as well as higher and higher framerates - something has to give, and computing fewer pixels totally makes sense to me. Thanks again for the talk, it was really easy to follow and had an excellent pace, not too fast and not too slow

Graphics Programming Conference 2025 presentation recordings are now online: youtube.com/watch?v=iRVLKi…

@KostasAAA SW raster of variable size triangles is a hard problem.

First stab at a compute shader rasteriser with albedo, uv and normal interpolation, clipping and depth testing. Works but it is 870x slower than the hardware rasterised version on the RTX 3080 mobile at 1080p.

Highly worthwhile read as always from Kostas!

I ported one of my radiance cascades shaders to @XorDev's new website and it's really fun how you can easily switch between passes and see how the algorithm converges in action. fragcoord.xyz/s/u4uhc9b9

Adaptive Resolution for AO: splits the screen into 2x2 tiles and traces at 0.25🟩, 0.5🟨, or 1.0🟥resolution based on geometry uniformity within a tile / disocclusion. Then uses 2-frame temporal upscaling to approach near full-res quality. #gamedev #indiedev #unity3d #raytracing

@idSoftwareTiago VRCS should be saving a ton of frame time! 😉

Trying to ski St Moritz Corviglia mountain ⛷️ Someone cranked fog strength to 11.