The recent surge of interest and commercial development in deep learning and artificial intelligence is impacting all areas of science and engineering where massive data can be harnessed to make autonomous decisions in real-time, ranging from consumer choice analytics and medical diagnosis to driverless cars. The key enabling factor that is fueling this revolution in pervasive computational intelligence is the seemingly limitless availability of data and computational power. New architectural advances in alternative non-von Neumann computing are needed to address design challenges faced in extending the relentless pace of exponential growth to sustain continued advances. This session highlights such new and emerging developments in large-scale accelerator hardware and software for deep learning and adaptive computational intelligence, including testimony by three companies in the IT marketplace: Intel, Facebook, and DeePhi.