Modders report DLSS 5 yields up to 90% performance boost on unsupported games
Modders have extracted the DLSS 5 Neural Rendering build 310.9.16.0 and applied it to games that Nvidia does not officially support, such as The Witcher 3. Their tests show performance improvements of up to 90% compared with the native DLSS 5 implementation, and about 40% faster than the previous NR model.
Key points
- Modders extracted DLSS 5 build 310.9.16.0 and ran it on games like The Witcher 3, seeing up to 90% performance gain.
- Nvidia says the new neural rendering model is 24% smaller, uses 20% less VRAM at 4K, and runs up to 2× faster.
- The update is about 40% faster than the previous NR model and includes both old (Preset A) and new (Preset B) versions.
Nvidia says the new neural rendering model is 24% smaller, consumes 20% less VRAM at 4K resolution, and processes its workload up to two times faster. The update ships with two presets – the older model (Preset A) and the newer, more efficient model (Preset B). Modders also adapted the runtime to work on RTX 20‑series GPUs using tools like OptiScaler, even though Nvidia only plans an official DLSS 5 rollout for RTX 40‑series cards.
Model page: DLSS 5 →
The story so far
3 episodes →- Modders report DLSS 5 yields up to 90% performance boost on unsupported gamesthis story
Official DLSS 5 running on unsupported games delivers up to 90% better performance than the leaked version
Tom's Hardware · 18 October 2026
Loading the full article…
This text was published by Tom's Hardware and written by Jhet Borja. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Meta releases Muse Spark Open, touted as highest‑performing open‑weights model in US · 3 src
- Google unveils Gemini 4 Argon with 1 million-token output · 4 src
- DeepSeek releases DeepSeek-V4.1-Flash-1017, scores 77.0 on DeepSWE v1.1 and opens Search and Fetch APIs · 13 src
- DeepSeek releases V4.1 Flash 1017 gguf model with refusals cut to 4% · 1 src
- MIT researchers unveil multimodal AI model that links crystal structures to properties · 1 src
Comments
via GitHub Discussions