The slowing of Moore's Law requires innovations in other areas. In order to make computers run faster, we no longer have the free lunch given from transistor shrinks anymore. We must explore accelerators and domain-specific architectures.

I am interested in the areas of computer architecture, compilers, VLSI, CAD for advanced technology nodes. Currently I am working with Prof. Gennady Pekhimenko and the EcoSystems Group on my thesis, in the area of GPU architecture.