The 330 MHz observing system at the VLA is a potentially powerful survey system. It can map fields many degrees in size quickly and at high sensitivity. However imaging characteristics unique to this data pose excessive computational burdens on conventional mapping systems. Hence while data acquisition is quick and efficient, data reduction has been difficult and slow. Here we describe how powerful new scalable processing algorithms have been used to generate the first full resolution 330 MHz images, demonstrating also that lower resolution survey work is now a tractable problem.