: Models were tailored for specific GPUs or CPUs (FP16 vs. FP32).
The release of version 5.5 wasn't just a minor patch; it introduced a brand-new AI engine. This architecture allowed the software to run on modern hardware. By moving to a more efficient processing pipeline, Topaz Labs addressed the #1 complaint of its users: the agonizingly slow processing times. Why the "Models" Files Mattered 2021-06-21_15.12-gigapixel-5.5-models-win.part2...
The 5.5 era was a turning point. It marked the moment AI upscaling moved from a "slow-but-cool" experimental tool to a high-speed professional workflow. While some users initially missed the simplicity of a single "all-in-one" installer, the modular model system allowed for much more frequent and specialized updates to individual AI models. : Models were tailored for specific GPUs or CPUs (FP16 vs