[en] Full text: L-H transition and pedestal buildup physics are at the top of the priority list for ITER physics research. Understanding the turbulence transport in the edge pedestal region is the key first step. The ion temperature gradient (ITG) driven turbulence is the most robust and baseline phenomenon which should be understood prior to the addition of other turbulences. However, due to the existence of the magnetic separatrix, first principles numerical studies of the 1) pedestal area have been difficult. Using our new full-f global gyrokinetic code XGC1, we have succeeded in the simulation of ITG turbulence in the edge pedestal region across the magnetic separatrix. The outer simulation boundary is the grounded material wall and the inner boundary can be anywhere in the core plasma. Unlike what the local theories have been indicating, we find that the nonlocal ITG turbulence exists across the whole density pedestal. Turbulence is compressed by the background free energy envelope whose scale length is about the same as the meso scale turbulence activities. Due to the interaction of ITG with background plasma the turbulence does not grow linearly, as is usually observed in a conventional delta-f simulation. Instead of the radial streamers, anisotropic eddies develop from the beginning of the ITG mode growth. Another significant finding from this research is the propagation of turbulence front from the pedestal region into the core region, affecting the self-organization property and marginality of the core plasma. We find evidences that the pedestal turbulence property is nonlocally connected to the core transport property. Other important physics phenomena, such as the heat flux profile, will also be reported. [Work performed under US DOE support to SciDAC Proto-FSP Center for Plasma Edge Simulation. XGC1 code is a collaborative code in CPES, technically managed by S. Ku of New York University] (author)