The envisioned next-generation CMB experiment, the so-called CMB S4, requires a large number of detectors (∼500,000) to achieve the necessary sensitivity. This is two orders of magnitude greater than current observing instruments, and an order of magnitude greater than new instruments under construction. A technology leap is necessary to achieve such large throughput in both fabrication and quality control. We are focused on this technology jump in two main fronts, taking advantage of experience and capabilities at LBNL. The first is the industrialization of the superconducting CMB detector wafer production. We rely on significant experience at the Micro-Systems Lab (MSL), which successfully industrialized the mass production of LBNL fully depleted CCDs. The CMB technology is based on superconducting transition-edge sensors (TES), which were pioneered by Adrian Lee’s group at UCB and are now the standard for the CMB field. A needed development that goes hand-in-hand with mass production is a factor ten improvement of frequency-domain multiplexing (fMUX) readout. This technology was also pioneered in Berkeley by H. Spieler, M. Dobbs, and A. Lee and is now the standard for CMB readout. We will pursue an order of magnitude increase in multiplexing by increasing the carrier frequency to 50 MHz. For this we are developing high-quality LC resonators made lithographically on a wafer substrate at the MSL. In addition to these two major developments, we are also developing a new, multi-mode detector technology. A multi-mode detector would achieve sensitivity improvement without increasing the number of channels and thus isan orthogonalto the development aiming the increase of the number of channels.