What you want is the Sweet Sixteen fader bank (which I seem to be bringing up a lot lately). It's an adaptation of this open source module concept, but it also has inputs in addition to outputs, allowing it to do the purely CV-based attenuation you want.
Chace and alexbreugelmans