New Nvidia DLSS rival promises faster frame generation on older AMD, Intel and GeForce GPUs

Nvidia DLSS has a new challenger in the form of Apex Neural Rendering (ANR), a suite of AI-powered rendering technologies developed by Apex Compute. ANR promises compatibility with graphics cards from Nvidia, AMD and Intel, including some fairly old models. ANR Frame Generation (ANR FG) supports Nvidia GeForce GTX 16-series and RTX 20-series GPUs or newer, AMD Radeon RX 5000-series GPUs or newer, and Intel Arc A-series and B-series graphics cards. The Steam Deck is also supported. On the software side, the currently available beta supports Windows 10/11, Linux and SteamOS through Proton.

Apex Neural Rendering consists of three technologies: frame generation, super-resolution upscaling and ray reconstruction. At the time of writing, only frame generation is available in beta. The other two components are expected to follow suit later.

Apex Compute has also published some interesting performance figures. On a GeForce RTX 5090 running at 4K, ANR reportedly takes 1.01 milliseconds to generate a frame in 2x mode, compared to 1.26 milliseconds for Nvidia DLSS Frame Generation. At 4x, those figures drop to 0.61 and 0.82 milliseconds, respectively. ANR also reportedly consumes just 214 MB of GPU memory, compared to 562 MB for DLSS. If accurate, ANR effectively highlights one of the primary pain points of frame generation: added latency.

The company claims its solution produces better image quality, too, and has published several side-by-side comparisons, including scenes from The Last of Us Part II. Of course, these are first-party benchmarks, and independent testing will be necessary to establish whether ANR can genuinely outperform DLSS in actual games. Faster frame generation alone does not necessarily translate into better overall performance or lower input latency.

The ANR FG beta is available on GitHub and currently relies on OptiScaler, which we previously covered. Installation requires modifying individual game directories, and it is not as simple as moving a few sliders around in a game’s settings. Compatibility will therefore vary between games.