GOSPEL-2 aims at exploring, implementation and demonstrating a novel concept for generation of ultra-broadband arbitrary waveforms in the optical or the THz frequency range, thus overcoming the limitations of current digital-to-analogue converters (DAC). The concept allows for massive spectral parallelization of conventional DAC interfaces by phase-correct interleaving of optical waveforms that are modulated onto the tones of an optical frequency comb. Building upon newly conceived schemes from the first funding period, we now aim at demonstrating waveform generation with an overall bandwidth of more than 300 GHz and at realizing an integrated signal generator that combines advanced photonic integrated circuits with RF electronics in a chip-scale module. The main conceptual challenge of optical arbitrary waveform generation (OAWG) systems is the phase control required for precise coherent combination of different tributary signals. In the first phase of the project, we conceived a novel concept to realize this phase control and successfully demonstrated the synthesis of a 200 GHz-wide optical waveform from two tributary signals. In the next funding period, we now wish to expand on this idea and utilize it to implement an integrated signal-generator module. This will require advanced photonic integrated circuits that are seamlessly connected to RF driver circuitry through newly developed dense broadband RF interfaces. At the same time, we will increase the number of signal tributaries from two to four, offering an overall bandwidth in excess of 300 GHz. Within the project, we shall develop an advanced model to quantitatively predict the performance of our system, design and implement the underling photonic integrated circuits and RF structures, implement a hybrid integrated photonic-electronic signal generator engine, and experimentally demonstrate the viability of the scheme for generation of ultra-broadband optical waveforms. We will further explore down-conversion of the optical waveforms to the THz frequency range through ultra-fast uni-traveling-carrier (UTC) photodiodes in the framework of a collaboration with an internationally leading group in this area. Based on an experimentally verified quantitative model of our system, we shall finally analyze the bandwidth-scalability of our scheme and evaluate the potential of synthesizing waveforms with THz bandwidths. We expect that the contents of GOSPEL-2 will be of high relevance both for the field of ultra-broadband photonic-electronic waveform synthesis and for the field of optical communications, where grid-less software-defined optical transmitters might allow for dynamic bandwidth allocation.