Over the past three decades, urban planners have attempted to make cities more sustainable by espousing higher density urban design concepts such as Compact Cities, Walkable Communities, and New Urbanist developments. It has been argued by some urban planners that the per capita energy use and air pollution emissions in densely built cities are less than in their more sprawling less dense counterparts. However, as urban density increases, the ability for pollutants to be transported out of the urban area is inhibited. This complex interaction between various types of urban form and their potential energy use and air quality is poorly understood. Through this work, we are attempting to increase knowledge for how environment and urban form interact. Our hypothesis is that urban structures and layouts exist which can minimize energy use while also minimizing air pollution exposure. The purpose of this work will be to investigate this complex interaction for various types of urban structures and to develop a design strategy for optimizing urban form under a variety of constraints. Our approach will be to develop an extremely fast and inexpensive energy use and dispersion modeling tool for urban areas that builds on our previous work. The modeling system will utilize the unique computational parallelism afforded by graphics processing units (GPUs, that are regularly utilized in the video game industry), to run many simulations in an effort to train an optimization algorithm for determining optimal designs for urban structures and their layout. We will also utilize an interactive and immersive virtual environment to provide unprecedented understanding and refinement of the complex physical processes associated with the energy balance and pollutant dispersion in an urban setting. Here, we present the model development strategy along with preliminary design optimization results for air quality only.