Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Nvidia announced DLSS 5 on March 16, 2026, as a real-time neural-rendering system that combines conventional game rendering with generative AI. A day later, after viewers criticized preview footage for looking uncanny and overly artificial, Nvidia CEO Jensen Huang gave a blunt response at GTC 2026: “Well, first of all, they’re completely wrong.”
That answer addresses one part of the controversy—the belief that DLSS 5 removes developers’ artistic control. It does not settle the separate questions of whether the output looks good, whether it preserves a game’s intended art direction, or whether players will want it enabled. Those questions require public builds and independent testing.
What Jensen Huang said about DLSS 5
During a press Q&A at Nvidia’s GTC 2026 event on March 17, Tom’s Hardware editor-in-chief Paul Alcorn asked Huang about the negative reaction to DLSS 5. Huang opened with: “Well, first of all, they’re completely wrong.”
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchHuang’s technical explanation was that DLSS 5 does not simply replace a game’s artwork with uncontrolled AI-generated imagery. Nvidia says the system combines game-engine information—including controllable geometry, textures and other rendered data—with a generative AI model. Developers can then tune the model to fit a game’s artistic style.
#1 Best Overall
- AI Performance: 767 AI TOPS
- OC mode: 2632 MHz (OC mode)/ 2602 MHz (Default mode)
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Axial-tech fan design features a smaller fan hub that facilitates longer blades and a barrier ring that increases downward air pressure
- A 2.5-slot design maximizes compatibility and cooling efficiency for superior performance in small chassis
In Nvidia’s framing, that combination of 3D graphics and artificial intelligence is why the company calls DLSS 5 “neural rendering.” Huang’s argument was therefore about control: developers still decide what the game is supposed to look like, while the neural system helps produce lighting, materials and other visual details at runtime. Tom’s Hardware reported the Q&A and Huang’s response.
That is a narrower defense than the quote might suggest. It challenges the idea that DLSS 5 is an unconstrained image generator, but it does not prove that every generated result will be attractive, faithful or preferable to the original render.
What Nvidia announced on March 16
Nvidia describes DLSS 5 as a real-time neural-rendering model that “infuses pixels with photoreal lighting and materials.” The company presents it as a bridge between conventional, artist-controlled game rendering and the kind of visual effects traditionally associated with offline film production.
Free tools Windows power users keep installed
One-click scans. No signup required.
That makes DLSS 5 more ambitious than a conventional resolution-reconstruction feature. Earlier DLSS technologies primarily focused on reconstructing a higher-resolution image from a lower-resolution render, generating additional frames, or—on supported hardware—creating multiple frames between traditionally rendered ones.
DLSS 5 is positioned as a system that can influence the appearance of lighting, surfaces and materials as part of the rendering pipeline. Nvidia’s official description explicitly says it combines handcrafted rendering with generative AI. Calling it “not AI” would therefore be inaccurate. Calling it merely an image-generation filter would also miss the role of structured game data and developer controls.
Nvidia listed Bethesda, Capcom, Hotta Studio, NetEase, NCSOFT, S-Game, Tencent, Ubisoft and Warner Bros. Games among the companies supporting the technology. That indicates announced industry participation, not universal developer enthusiasm or proof that every integration will produce the same result. Nvidia’s DLSS 5 announcement and its newsroom release describe the technology and its intended role.
Rank #2
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5070 Ti
- Integrated with 16GB GDDR7 256bit memory interface
- PCIe 5.0
- WINDFORCE cooling system
Why gamers reacted so negatively
The backlash was driven primarily by what the preview footage appeared to show, rather than by a detailed rejection of Nvidia’s developer-tooling explanation.
Viewers particularly focused on character footage from Resident Evil Requiem, including Grace Ashcroft and Leon Kennedy. Critics described faces as uncanny, waxy, glossy or visually homogenized. Some said the images resembled “AI slop”—a criticism of the apparent aesthetic result, not a verified technical classification of the system.
Faces are especially revealing in this context. Small changes to skin, eyes, hair, expressions and facial structure can make a familiar character look synthetic even when the rest of a scene appears technically impressive. A system that improves apparent realism in lighting may still reduce fidelity to the character model or the game’s chosen visual style.
The fear of an “Nvidia look”
Another concern is homogenization. Critics worry that neural-rendering models could impose a recognizable visual signature across different games, even when developers supply the original assets and artistic guidance.
That outcome has not been established as an unavoidable technical fact. It is a reasonable concern about how a shared model might affect different art directions, particularly stylized games or titles that deliberately avoid photorealism. A photorealistic horror game, a cel-shaded action game and a painterly fantasy game may not benefit from the same interpretation of materials or lighting.
Control is not the same as approval
The philosophical criticism goes further than asking whether developers can adjust parameters. Players are also asking whether the final image still expresses deliberate, human-authored art direction—or whether an AI model is making visible creative decisions at runtime.
Rank #3
- Powered by the NVIDIA Blackwell architecture and DLSS 4. System Requirements: Minimum 850W PSU with 16-pin 12V-2x6 (12VHPWR) connector required. Verify before purchasing.
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability. Compatibility: 348mm (13.7") length, 3.6 slots, 4.3 lbs. Confirm case clearance and slot spacing. GPU bracket included.
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.6-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
Nvidia’s answer is that developers remain in control and can tune the system to a game’s style. That directly addresses ownership of the controls. It does not automatically answer whether the resulting image is faithful to the artist’s intent, whether it looks better in motion, or whether players prefer it to native rendering or conventional DLSS.
Nvidia’s defense versus the criticism
| Nvidia’s position | What critics are still asking |
|---|---|
| DLSS 5 uses game geometry, textures and other structured rendering information. | Does using those inputs prevent unwanted changes to faces, materials, animation and fine detail? |
| Developers can fine-tune the model for a game’s artistic style. | Does developer control guarantee an appealing result, or merely make the result adjustable? |
| The system combines handcrafted rendering with generative AI. | Will the AI component produce stable, faithful images during movement and across varied scenes? |
| Neural rendering can add photorealistic lighting and materials. | Is more photorealism desirable in a game whose art direction is intentionally stylized or less realistic? |
Both sides can therefore be right about different questions. Huang may be correct that DLSS 5 is constrained by game data and developer input. Gamers may still be correct that the demonstrated output looked undesirable. Technical controllability and visual appeal are not interchangeable claims.
How DLSS 5 differs from earlier DLSS
The DLSS name creates some of the confusion because many players associate it mainly with performance-focused reconstruction.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- DLSS Super Resolution reconstructs a higher-resolution image from a lower-resolution render.
- DLSS Frame Generation creates additional frames between traditionally rendered frames.
- DLSS Multi Frame Generation creates multiple additional frames on supported hardware.
- DLSS 5 is being presented as a broader neural-rendering system that can affect lighting, materials and the visual appearance of rendered pixels.
DLSS 5 is therefore not simply “another AI upscaler.” Nvidia is extending the brand into a more comprehensive rendering approach. That may ultimately make it more powerful, but it also raises a higher standard: players will judge not only frame rate and reconstruction quality, but whether the system respects the visual identity of each game.
Nvidia’s current RTX strategy already describes neural rendering as the integration of neural networks into the graphics pipeline. DLSS 5 appears to be a major expansion of that idea rather than a routine revision to image scaling. Related DLSS capabilities are described in Nvidia’s RTX 50-series materials and its DLSS 4.5 announcement.
Was the demo running on two RTX 5090s?
Reports from the GTC demonstrations said DLSS 5 was shown using two RTX 5090 graphics cards: one rendering the game and another handling the neural-rendering workload. That is important context, but it is not a confirmed consumer requirement.
Rank #4
- Powered by the NVIDIA Blackwell architecture and DLSS 4
- Powered by GeForce RTX 5060
- Integrated with 8GB GDDR7 128bit memory interface
- PCIe 5.0
- WINDFORCE cooling system
Conference demonstrations can use development hardware, unusually high settings or a configuration chosen to make a prototype work reliably. Nvidia may optimize the implementation, distribute the workload differently or establish different requirements before release.
Recommended Free Tools
Readers should not conclude that every DLSS 5 game will require two RTX 5090s. They should also not assume that every RTX 50-series card will deliver identical performance or image quality. Nvidia’s announced platform context points to GeForce RTX 50-series hardware, but the reviewed materials do not provide a complete final compatibility matrix. There is no confirmed basis here for claiming support on RTX 40-series or older cards.
When will DLSS 5 arrive?
Nvidia announced DLSS 5 on March 16, 2026, and said it would arrive in fall 2026. The cited announcement did not provide a firm launch date.
Actual availability is likely to depend on more than a graphics-card driver. Support may vary by game, developer integration, SDK version, driver release and GPU model. A game appearing in a technology demonstration does not necessarily mean its final public build will ship with identical settings or quality.
The practical buying advice is straightforward: do not purchase an RTX 50-series card solely for an untested future DLSS 5 promise. Buy one based on current performance and existing features—including current DLSS capabilities, ray tracing and the games you play. DLSS 5’s final requirements and real-world behavior should be judged after public software becomes available.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →What remains unknown
Before DLSS 5 can be evaluated properly, independent reviewers and players need answers to several questions:
Best Value
- Powered by the NVIDIA Blackwell architecture and DLSS 4 OC mode: 2640MHz/Default mode: 2610MHz (Boost Clock)
- Military-grade components deliver rock-solid power and longer lifespan for ultimate durability
- Protective PCB coating helps protect against short circuits caused by moisture, dust, or debris
- 3.125-slot design with massive fin array optimized for airflow from three Axial-tech fans
- Phase-change GPU thermal pad helps ensure optimal thermal performance and longevity, outlasting traditional thermal paste for graphics cards under heavy loads
- Which RTX 50-series models will support it, and will quality or performance vary by model?
- Will RTX 40-series or older GPUs receive any form of support?
- How much GPU time, VRAM and Tensor Core capacity will the neural pass require?
- Does neural processing add latency, and how does that interact with frame generation and latency-reduction technologies?
- How stable are faces, hair, particles, transparency, reflections, subtitles and UI elements during motion?
- Can developers tune the system by scene or choose different quality modes?
- Can players disable DLSS 5 and use native rendering, conventional DLSS or another fallback?
- Will final builds address the visual issues critics saw in the preview footage?
A proper comparison should include native resolution, conventional DLSS Super Resolution, earlier DLSS modes and—where relevant—ray-traced alternatives. A result that looks convincing in a carefully selected demo may behave differently in an open-world game, an older engine, a stylized title or a heavily modded installation.
Did Huang later soften his position?
Later coverage of an appearance on the Lex Fridman podcast described a more conciliatory tone. Huang reportedly said he understood why people reacted negatively and that he did not like “AI slop” himself, while maintaining that DLSS 5 was not intended to produce that result.
That follow-up does not erase his GTC response, but it helps distinguish the company’s intent from the audience’s reaction. The available secondary report supports a more empathetic explanation, not necessarily an apology. TechRadar reported the later comments.
The verdict: Huang answered the architecture question, not the taste question
Jensen Huang’s “completely wrong” remark was aimed at the idea that DLSS 5 takes away all artistic control and substitutes uncontrolled AI imagery for a game’s authored assets. Nvidia’s explanation supports a more constrained model: the system uses structured game data, developer guidance and generative AI together.
But that defense does not prove that the preview looked good, that the output will preserve every artist’s intent, or that players will prefer it. The backlash is about image quality, authenticity, aesthetics and trust as much as it is about implementation.
Until DLSS 5 reaches public games, the fairest conclusion is provisional: Nvidia may be right about how the system is designed, while gamers may still be right that the demonstrated results were unconvincing. The final judgment depends on independent comparisons, game-by-game behavior, performance, latency and the controls developers and players actually receive.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →
