A videoconferencing apparatus includes a multi-point (MP) conference application that enables the apparatus to combine and distribute audio and video signals received from a plurality of remote conference endpoints, thereby obviating the need to provide a separate multi-point control unit having hardware-based inverse multiplexers (IMUXs). The MP conference application is configured to generate, for each remote conference endpoint participating in a conference, discrete instances of a signal processing train by means of dynamically allocable IMUXs, each processing train including a communication process and audio/video/data codecs. The processed audio and video signals are subsequently conveyed to an audio mixer and video switching module for combination with locally-generated audio and video signals. The outputs of the audio mixer and video switching module are sent to each of the plurality of signal processing trains, which process the combined signals according to a transmit mode for distribution to the remote endpoints over the network.