Monday, March 16th 2026

NVIDIA Unveils DLSS 5 with Real-Time Neural Rendering

NVIDIA today unveiled NVIDIA DLSS 5, the company's most significant breakthrough in computer graphics since the debut of real-time ray tracing in 2018. DLSS 5 introduces a real-time neural rendering model that infuses pixels with photoreal lighting and materials. Bridging the divide between rendering and reality, DLSS 5 empowers game developers to deliver a new level of photoreal computer graphics previously only achieved in Hollywood visual effects.

"Twenty-five years after NVIDIA invented the programmable shader, we are reinventing computer graphics once again," said Jensen Huang, founder and CEO of NVIDIA. "DLSS 5 is the GPT moment for graphics—blending hand-crafted rendering with generative AI to deliver a dramatic leap in visual realism while preserving the control artists need for creative expression."
Bridging The Cinematic Gap
Since the dawn of GeForce, NVIDIA has strived to deliver the graphics horsepower required for game developers to create incredible, realistic worlds—where lighting, reflections and shadows obey the laws of nature.

From programmable shaders with GeForce 3 in 2001, to CUDA with GeForce 8800 GTX in 2006, to real-time ray tracing with GeForce RTX 2080 Ti in 2018, to path tracing and neural shaders with GeForce RTX 5090 in 2025, NVIDIA has delivered major architectural innovations and a massive 375,000x increase in compute to meet this challenge.

However, the rendering horsepower available to a 16-millisecond game frame remains a tiny fraction of that available to a photoreal Hollywood VFX frame, which can take minutes to hours to render. Real-time rendering cannot bridge the gap to photorealism through brute force alone.

NVIDIA DLSS was released in 2018 as an AI technology to boost performance, first by upscaling resolution, and then by generating entirely new frames. It has been integrated in over 750 games, becoming a gold standard for the industry. Launched at CES this year, DLSS 4.5 uses AI to draw 23 out of every 24 pixels seen on the screen. Today, DLSS is evolving beyond performance to transform visual fidelity in games.

Video AI models have rapidly learned to generate photoreal pixels, but they run offline, are difficult to precisely control and often lack predictability, with every new prompt generating bespoke content. For games, pixels must be deterministic, delivered in real time and tightly grounded in the game developer's 3D world and artistic intent.

DLSS 5 takes a game's color and motion vectors for each frame as input, and uses an AI model to infuse the scene with photoreal lighting and materials that are anchored to source 3D content and consistent from frame to frame. DLSS 5 runs in real time at up to 4K resolution for smooth, interactive gameplay.

The AI model is trained end to end to understand complex scene semantics such as characters, hair, fabric and translucent skin, along with environmental lighting conditions like front-lit, back-lit or overcast—all by analyzing a single frame. DLSS 5 then uses its deep understanding to generate visually precise images that handle complex elements such as subsurface scattering on skin, the delicate sheen of fabric and light-material interactions on hair, all while retaining the structure and semantics of the original scene.

DLSS 5 provides game developers with detailed controls for intensity, color grading and masking, so artists can determine where and how enhancements are applied to maintain each game's unique aesthetic. Integration is seamless, using the same NVIDIA Streamline framework used by existing DLSS and NVIDIA Reflex technologies.

Availability & Game Developer Support
DLSS 5 will be supported by the industry's biggest publishers and game developers, including Bethesda, CAPCOM, Hotta Studio, NetEase, NCSOFT, S-GAME, Tencent, Ubisoft and Warner Bros. Games.

"Bethesda has such a rich history pushing graphics with NVIDIA, going all the way back to Morrowind, with that incredible water," said Todd Howard, studio head and executive producer at Bethesda Game Studios. "When NVIDIA showed us DLSS 5 and we got it running in Starfield, it was amazing how it brought it to life. We've played it. We can't wait for all of you to do so as well."

"At CAPCOM, we strive to create experiences that feel cinematic, compelling and deeply believable—where every shadow, texture and ray of light is crafted with intention to enhance atmosphere and emotional impact," said Jun Takeuchi, executive producer and executive corporate officer at CAPCOM. "DLSS 5 represents another important step in pushing visual fidelity forward, helping players become even more immersed in the world of Resident Evil."

"Immersion is about making the world feel real. DLSS 5 is a real step towards that goal," said Charlie Guillemot, co-CEO of Vantage Studios. "The way it renders lighting, materials and characters changes what we can promise to players. On Assassin's Creed Shadows, it's letting us build the kind of worlds we've always wanted to."

DLSS 5 will come to games including AION 2, Assassin's Creed Shadows, Black State, CINDER CITY, Delta Force, Hogwarts Legacy, Justice, NARAKA: BLADEPOINT, NTE: Neverness to Everness, Phantom Blade Zero, Resident Evil Requiem, Sea of Remnants, Starfield, The Elder Scrolls IV: Oblivion Remastered, Where Winds Meet, and more.

DLSS 5 will arrive this Fall. See the first preview this week at GTC, and below, you can check out several examples of NVIDIA DLSS 5 in Resident Evil Requiem, EA SPORTS FC, Starfield, Hogwarts Legacy, and the NVIDIA Zorah tech demo.

Resident Evil Requiem
EA SPORTS FC
Starfield
Hogwarts Legacy
NVIDIA Zorah Tech Demo
Source: NVIDIA
Add your own comment

305 Comments on NVIDIA Unveils DLSS 5 with Real-Time Neural Rendering

#51
Blueberries
A UFO could crash in my backyard and it still wouldn't be as mind-blowing as the DLSS5 Demo I just watched
Posted on Reply
#52
LastDudeALive
On the bright side, this may stop developers from pushing even more polygons into their games in the quest for more "realism." Instead of relying on 1 trillion polygons to have micron-accurate 3D models and Lumen tanking performance on a 5090, they can have less detail and let DLSS do the work of making it more accurate.
Posted on Reply
#54
gffermari
Impressive. I like it but the developers need to have full control of what DLSS 5 produces.

What's more interesting is that in AC Shadows, DLSS 5 transforms the lighting in the environment completely although the game uses a GI solution.
The question is then...why should the developer of a future game even produce a raster, RT or PT version of the game. They could just let the lighting simulation to be entirely handled by DLSS 5 and call it a day.
Posted on Reply
#55
oxrufiioxo
RejZoRIt's impressive, but at the same time I have issues with this (and I'm generally very favorable and also forgiving to upscaling tech). The Resident Evil Requiem example it looks like it's entirely different woman being showcased. Which puts artistic design intention of developer to question if graphic card then decides to just render something entirely different in the end.
Yeah, I've went back and fourth with these two screens probably 50 times and I think honestly she looks different mostly due to lighting at least in a cutscene which has a much higher quality Grace model vs the in game one. I think if I had never seen the original game and this neural rendered version was my first look at it my view would be slightly more positive but right now I'm just neutral thinking this could go either way good or bad.

I do like the way it handles eyes and can see it handling skin really well if the developer is given full control.

Posted on Reply
#56
RejZoR
LastDudeALiveOn the bright side, this may stop developers from pushing even more polygons into their games in the quest for more "realism." Instead of relying on 1 trillion polygons to have micron-accurate 3D models and Lumen tanking performance on a 5090, they can have less detail and let DLSS do the work of making it more accurate.
Well, if developers used existing tech right, we wouldn't need any of this. I'm still salty how idiotic EVERYONE used Tessellation. Everyone just slapped it on low poly models and cranked it to 11. Games churned out idiotic amounts of polygons and usually on shit that didn't matter. So we still had insanely blocky railings and guns had octagonal scopes on them, but they pushed 30 million polygons on some random shit in the world.

Tessellation should really be used in reverse as full replacement for old LOD. It should be performance saving feature with seamless downgrading of models that don't flip radically in visuals past certain distance. That's what tessellation was really good at as models just gradually and slowly changed where LOD flipped between usually 3 different models based on distance. We never got that, but we got drastically shit performance in many games because basically everyone used it wrong.
Posted on Reply
#57
sbacc
Digital Foundry saying in their YouTube video that it's the same geometry and the same texture assets, only lighting is changed, enhanced with realism and saying this while in the background the RE demo show obvious changes in the face of the female character in the street scene, just look at her lips !

At this point either DF is blind or incompetent or just an unofficial PR arm of Nvidia...

edit :
Posted on Reply
#58
LastDudeALive
vlad.coolishSoon... All NPC in all games.
:D
Hold up, maybe Nvidia is onto something...
Posted on Reply
#59
Fatloader
The environmental changes are quite impressive where you can notice them, but the faces look terrible. They should've waited to showcase this until it was closer to release and looked better.
Posted on Reply
#60
Onasi
It’s an impressive bit of technology, but, as presented on the slides, it just kind of looks like ass. This is going to be a fairly old-school reference, but it has that “FakeFactory Cinematic Mod” kind of look. It wants to do a very specific thing and that thing is “filmic” and “photorealistic” and “lifelike” and… and I am already bored and want to go to sleep. There are people who would appreciate such an aesthetic, but it looks incredibly trite to me. But, as mentioned before, this is just slapped on top of existing games. It would be curious to see one developed specifically to take advantage of this and see how that fares.
Posted on Reply
#61
Imsochobo
this feels like a bad april fools joke...
Posted on Reply
#63
RejZoR
sbaccDigital Foundry saying in their YouTube video that it's the same geometry and the same texture assets, only lighting is changed, enhanced with realism and saying this while in the background the RE demo show obvious changes in the face of the female character in the street scene, just look at her lips !

At this point either DF is blind or incompetent or just an unofficial PR arm of Nvidia...

edit :
See, the thing is, all the assets may be original. But whatever NVIDIA trained their models on think the lips should look like that and it slapped that kind of texture on her face. And same goes for eyes shadow and even eyes themselves. Because if you look at the original, there is NO WAY a simple "upscale and enhance" would ever do that. The Ai logic literally SWAPPED the rendered "assets".

I've done a lot of image processing and upscaling with Ai tools, the kind that doesn't just smear bullshit all over images to reconstruct them, it actually very faithfully enhances details. This is not it. This is generative level online services where you throw in photo of a woman, lets say Emily Rudd since she's popular recently, you describe it to the service in a prompt and tell the service to make her more sexy. And then the Ai spits out a generated visuals of what it thinks Emily Rudd is suppose to look like in a hotter version. This is what DLSS is.

It doesn't alter assets, they are still the same on your drive, it straight up swaps them during render process using generative Ai. Which is why I said before I have issues with it. Because this isn't artistic design of a game developer anymore, this is artistic design of NVIDIA. Also why I didn't have issues with DLSS or FSR as a function. It just really upscaled and enhanced visuals and they pretty much entirely remained faithful to design developers intended. This is NOTHING like that. This is ReShade level of altering. Difference between DLSS 5 and ReShade is that ReShade is entirely optional and doesn't impede performance where turning DLSS off to not have this crap smeared over models will mean you'll suffer with worse framerate. Unless DLSS 5 will be able to exist without this generative bs. Or downgrade to DLSS 4.x.

I don't know, I have mixed feelings about this. It's impressive, but not for reasons one would like to compliment it. This would be wildly impressive if it was "Realtime Remaster" for old games, but those don't support DLSS anyway so I don't know.
Posted on Reply
#64
Visible Noise
AMD feature set remains five years behind Nvidia. Time to start the RDNA 6 hype train.
Posted on Reply
#65
rv8000
More slop incoming.

In several of the image comparisons they’re drastically altering the lighting and textures (not talking about the character models, which is probably their main goal). In other words , they’re destroying artistic intent and the work of the designers and developers.

I can’t wait to see all the resulting artifacts and model glitching down the road, which will just get glossed over when DLSS 6 gets rammed down the pipeline.

Sigh…
Posted on Reply
#66
Vya Domus
sbaccAt this point either DF is blind or incompetent or just an unofficial PR arm of Nvidia...
They are hardcore Nvidia shills for sure but also pretty incompetent, worst part about it is that studios actually listen to them so whatever idiotic opinion they have on something ends up being taken as matter of fact.
Posted on Reply
#67
oxrufiioxo
YouseethisYou really need to read the FAQ :roll: :

www.nvidia.com/en-us/geforce/forums/rtx-technology-dlss-dxr/37/583738/dlss-5-faq/

This thing run on 2x5090. They will optimize it they say but will it reach a point that it will be able to run on a single 16gb card? SLI IS BACK WITH A VENGANCE! We know it works with 2 cards, will this ability be removed on release?

Nvidia claims that this thing has controllability at least.
Wouldn't be surprised if the 5090ti released at the same time with the slogan "first consumer gpu that can handle DLSS5" :laugh: :toast:
Visible Noise:laugh: AMD feature set remains five years behind Nvidia. Time to start the RDNA 6 hype train.
FSR Diamond... Next gen only. :laugh:
Posted on Reply
#68
sbacc
RejZoRSee, the thing is, all the assets may be original. But whatever NVIDIA trained their models on think the lips should look like that and it slapped that kind of texture on her face. And same goes for eyes shadow and even eyes themselves. Because if you look at the original, there is NO WAY a simple "upscale and enhance" would ever do that. The Ai logic literally SWAPPED the rendered "assets".

I've done a lot of image processing and upscaling with Ai tools, the kind that doesn't just smear bullshit all over images to reconstruct them, it actually very faithfully enhances details. This is not it. This is generative level online services where you throw in photo of a woman, lets say Emily Rudd since she's popular recently, you describe it to the service in a prompt and tell the service to make her more sexy. And then the Ai spits out a generated visuals of what it thinks Emily Rudd is suppose to look like in a hotter version. This is what DLSS is.

It doesn't alter assets, they are still the same on your drive, it straight up swaps them during render process using generative Ai. Which is why I said before I have issues with it. Because this isn't artistic design of a game developer anymore, this is artistic design of NVIDIA. Also why I didn't have issues with DLSS or FSR as a function. It just really upscaled and enhanced visuals and they pretty much entirely remained faithful to design developers intended. This is NOTHING like that. This is ReShade level of altering. Difference between DLSS 5 and ReShade is that ReShade is entirely optional and doesn't impede performance where turning DLSS off to not have this crap smeared over models will mean you'll suffer with worse framerate. Unless DLSS 5 will be able to exist without this generative bs. Or downgrade to DLSS 4.x.

I don't know, I have mixed feelings about this. It's impressive, but not for reasons one would like to compliment it. This would be wildly impressive if it was "Realtime Remaster" for old games, but those don't support DLSS anyway so I don't know.
I know, but in the end the picture is changed beyond a simple lighting pass like you have with a ReShade AO, so telling everything is the same except lighting is just wrong...

This isn't a dodge/burning on a neutral gray layer overlayed on top of the OG picture, this is AI redraw
Posted on Reply
#69
lambda
NinWeird. I don't think these pictures are doing Nvidia a favour at all. The characters' faces look completely different from their actual designs. The ones in Resident Evil and Hogwarts Legacy especially. All these pictures have that GenAI feeling to them--they look beautiful and impressive, but they don't look right. Maybe in some top-down games I'd turn this on, but in games where character designs are important, I'd give this a wide berth.

On the first EA Sports picture, the cloth of Van Dijk's shirt even folds differently. Not the way the game natively renders it. It looks like DLSS is drawing on top of it with AI. ([Edit] Oh, it is. I should have read the text better.) Just like on the characters' faces, in fact. also note how Grace gets fuller lips in the screenshot where she's standing in the street, but not in the screenshot with the close-up of her face. Awful.
Agreed, the side by side comparisons aren't doing them a favour. All of them have this weird excessive polished and distinctively fake look to them.

With that said, it'll improve over the years with more training data as more games integrate such technologies in them.

I also hope we get proper hardware for running this stuff locally at affordable price points.
Posted on Reply
#70
Kinsa83
Thank you whoever mentioned the fact the dlss is doing more than just adding light cause I was sitting there pondering why do a good portion of the characters look less angry to me with the feature turned on. Example Leon. DLSS off his expression reads like "gods damnit, this bs again." In an irritated, but accepted reality way. With dlss on his face expresses a calmer "well this is the job, time to do it." Is the lighting really impacting how im interpreting the message of the moment that much? Light is a visual representation of hope and other positive things in art. I dont like it. This is going to alter artist intention.

Now lets think about people who cant tell the difference between something fake and something real. We already having people becoming more delusional from AI use with just conversing with text. Now imagine people interacting with something that is photo realistic and more peoples touch with reality will be blurred even more. People do have to make choices for themselves and they have control of what settings to use or not. But lets face it half of adults out in the world dont want to critically think. Im still waiting for people to figure out just because they can do something doesnt mean they should do it.
Posted on Reply
#71
Visible Noise
ChomiqF#ck off with this, totally ruining creative vision of actual artists working hard on creating these games.
Devs are ruining their own creative vision?

How exactly does that work??
Posted on Reply
#72
remekra
Just to start off, I have 5080 before anyone says that if AMD would release something like that I would be praising it.
But holy, wtf is this? It looks uncanny, extremely uncanny. It's basically like running Stable Diffusion at 60 fps. Impressive that they can do that, but it just looks bad. Like you would apply youtube shorts AI generated slop filter on every frame. Like there would be a little gremlin just typing into chatgpt "make this image look ultra realistic" at 60fps. I don't care if lightning is physcially accurate to 100 decimal point. It looks fake. AI generated fake.
If that's the future of graphics then I'm out.
Posted on Reply
#73
Guwapo77
I see this different than most, I think the characters look great with DLSS 5 On. The problem I have is the environment...everything brighter and it feels like it removes the atmospheric effects.
Posted on Reply
#74
Visible Noise
sbaccDigital Foundry saying in their YouTube video that it's the same geometry and the same texture assets, only lighting is changed, enhanced with realism and saying this while in the background the RE demo show obvious changes in the face of the female character in the street scene, just look at her lips !

At this point either DF is blind or incompetent or just an unofficial PR arm of Nvidia...
Imagine that, changing lighting - such as self shadowing - could make a face look different!
Posted on Reply
#75
Bigshrimp
Looks like generative AI was used, and not just lighting changes. The uncanny valley issue seems to still be there. It all looks strange to me.

They probably trained all of this AI on Jersey Shore and The Kardashians episodes, lol...
Posted on Reply
Add your own comment
Nov 20th, 2026 15:45 PST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts