During CES 2024, NVIDIA’s eagerly anticipated DLSS 3 technology faced challenging scrutiny. Initial reports suggest that the performance gains promised by the new tech may not be as impressive as users might have hoped. For those hoping to achieve at least a 40 frames per second (FPS) native framerate, which is critical for maintaining low input latency, the average improvement reportedly drops to 81%. This reduction in performance improvement is particularly noteworthy considering AMD’s recommendation of a native 60 FPS benchmark for their similar tech solutions, not to mention that expert reviewers such as the team at Hardware Unboxed suggest aiming for double that figure.
The heart of the issue with DLSS 3 appears to lie in its computational demands. When analyzing frame times, which measure the milliseconds it takes to render each frame, as opposed to frame rates (frames rendered per second), it becomes evident that NVIDIA’s RTX 4080 Super, while faster than the RTX 3080 Ti, does not achieve the performance uplift that might be expected when DLSS 3-generated frames are factored in. In essence, the gap between the actual time the RTX 4080 Super takes to render frames with DLSS 3 compared to the predicted 41% improvement over the RTX 3080 Ti reveals the additional time DLSS 3 requires to process. The specific figure reported is an additional 5.32 milliseconds per frame.
This revelation has sparked conversations among tech enthusiasts and industry watchers alike about the true cost of improved graphics fidelity through AI-enhanced rendering techniques. With NVIDIA’s DLSS 3 at the epicenter of these discussions at CES 2024, it seems that the technology must overcome some obstacles to meet and exceed consumer expectations fully. The tech community will undoubtedly be looking forward to more updates and optimizations as NVIDIA responds to the feedback from this significant industry event.
