Initial Impressions and Testing Methodology
The advanced upscaling technology, DLSS 5, has generated considerable anticipation since its announcement. While official media provided curated visual examples, testing the technology in real-world gameplay was necessary to gauge its true capabilities. Using the community-developed DLSS Swapper mod, the author tested DLSS 5 across multiple titles, finding the technology impressive but complex. The author noted that while DLSS 5 offers significant visual enhancements, it simultaneously challenges the core principles of upscaling technology, specifically regarding visual integrity and performance.
Test Bench and Game Selection
The testing was conducted on a system powered by a Ryzen 5 5600X processor and 16GB of DDR4 (3200 MHz) memory. The primary GPU used was an RTX 4070 equipped with 12GB of VRAM, providing ample resources for testing at both 1080p and 1440p resolutions. The author tested games including Cyberpunk 2077, Red Dead Redemption 2, and Onimusha: Way of the Sword. Additional titles evaluated were Monster Hunter Wilds, Elden Ring Nightreign, and Dragon’s Dogma 2. All games were installed freshly on a Gen3 NVMe SSD, utilizing the DLSS Swapper mod without other modifications.
When testing with an RTX 4060 (8GB VRAM), the author observed that the 8GB capacity became a significant limiting factor when running demanding games like Onimusha and Cyberpunk 2077 at 1440p, although the 4060 was surprisingly capable of running DLSS 5 despite these limitations.
Visual Impact and Performance Metrics
Visually, DLSS 5 is striking when operational. It substantially improves a game’s lighting and global illumination (GI), and reflections, drawing comparisons to the look achieved by UE5’s Lumen GI system. This enhanced lighting significantly impacts character models, making them appear more “clean” and artificial, yet still integrated within the game environment. However, the author cautions that this overhaul often diminishes the unique aesthetic and distinct art style of the original games.
Conversely, the most notable finding during the tests was the performance impact. Contrary to the technology’s intended purpose of boosting framerates, running DLSS 5 caused a substantial reduction in performance across all tested titles. The author noted that in every scenario, DLSS 4/4.5 utilizing the Quality preset provided a considerable performance advantage over the same preset when running with DLSS 5.
Performance Comparison Data (Average FPS)
The following metrics summarize the average framerates recorded during the testing process:
| Games Tested | Avg FPS 1080p (with DLSS 5) | Avg FPS 1440p (with DLSS 5) | Avg FPS 1080p (with DLSS 4 Quality) | Avg FPS 1440p (with DLSS 4 Quality) |
|---|---|---|---|---|
| Cyberpunk 2077 | 67 FPS (Ultra + No RT) | 54 FPS (Ultra + No RT) | 82 FPS (Ultra + No RT) | 73 FPS (Ultra + No RT) |
| Red Dead Redemption 2 | 62 FPS | 57 FPS | 88 FPS | 74 FPS |
| Onimusha: Way of the Sword | 70 FPS | 62 FPS | 80 FPS | 72 FPS |
Overall Assessment
In summary, while DLSS 5 showcases genuinely impressive advancements in lighting, the current implementation feels highly experimental. The technology excels at overhauling a game’s visuals but does so at the expense of performance and the original art style. The author concluded that the performance degradation outweighs the visual upgrades, suggesting that DLSS 5 does not currently replace the established performance and visual fidelity of DLSS 4/4.5.
