Imagine a “Commodore 64 Ultimate+”, built around either FPGA or high-accuracy emulation, but with a layer of AI and DSP enhancement on top.
Graphics: Every frame of 320×200 pixel art could be automatically upscaled using modern AI models trained on retro visuals. Instead of blocky sprites, you’d see smooth, detailed artwork that still preserves the classic C64 style. (Think ESRGAN, Real-ESRGAN, or even diffusion-based style transfer.)
Sound: The SID chip output could run through a real-time audio enhancer — stereo widening, reverb, or even neural resynthesis that reinterprets the waveforms as realistic instruments while keeping the iconic melodies intact.
Architecture: The C64 core would run normally (FPGA or emulator), with the enhanced audio/video handled by a separate processor or GPU and output over HDMI. Latency could be kept low enough for real gameplay. It would still play all the original games and demos — just better by default, like a retro console that’s been given a modern polish without losing its soul.
You could even toggle between “authentic” and “enhanced” modes on the fly.
So basically: a C64 that reimagines what those games wanted to look and sound like, using today’s tech.
Would you want something like this?