Recent developments have incorporated the WaveWatch III (WW3) spectral wave model into the fully coupled Energy Exascale Earth System Model (E3SM). However, WW3 is currently very computationally expensive to run at the spatial resolutions used by E3SM on climate time scales. This study presents ongoing work focused to testing the performance of WW3 for use in global climate-scale simulations. We present several avenues currently being explored to improve the computational performance of WW3 on climate time scales such as utilizing i) unstructured, regionally refined mesh grids ii) GPUs to calculate the wave action source terms iii) domain decomposition parallelization methods and iv) an implicit numerical scheme. These improvements to the efficiency of WW3 allow us to explore the implications of waves in a fully coupled climate system as well as the critical interactions of waves and sea ice in polar regions.