Low FPS in Eglantine

People ask how to raise Eglantine to 120 FPS, and a second group says frame generation does nothing. The switches file is the project’s own warning label. EDEN_NX2_SPLIT at 2 is described as moving GPU method and macro work to a background thread, with a claim of up to 54 percent more FPS on its own. EDEN_NX2_FENCE_BATCH at 1 is the frame-time smoother. EDEN_NX2_BUFFER_DOWNLOADS forced on is described as a 38 to 63 percent drop, and the file says to leave it off. EDEN_NX2_FRAMES_AHEAD at 2 cuts spikes and adds about one or two frames of input latency. EDEN_NX2_NO_REACTIVE is a Mario Kart experiment that one note drops from 29 to 24.

Resolution scale is the other lever that shows up in Mario Kart reports: 0.5x on a weak GPU, and handheld versus docked on an iGPU. Persistent shader cache and the Vulkan pipeline cache are handbook toggles, not forks. Frame generation, including the FSR and LSFG experiments on the ZendNevs builds, is something I would only discuss after the base frame rate without it is written down. The community note is to avoid it when the base rate is under about 50, because the counter lies and game speed on some titles follows the real frame rate. A 30 FPS cap is a fix for double-speed games, not a failure.

If you report this, include the base FPS with frame generation off, the build, the CPU, the GPU and the resolution scale. A screenshot of a generated 120 does not answer the question.