A blueprint for producing scalable digital graphene electronics has remained elusive. Current methods to produce semiconducting-metallic graphene networks all suffer from either stringent lithographic demands that prevent reproducibility, process-induced disorder in the graphene, or scalability issues. Using angle resolved photoemission, we have discovered a unique one-dimensional metallic-semiconducting-metallic junction made entirely from graphene, and produced without chemical functionalization or finite size patterning. The junction is produced by taking advantage of the inherent, atomically ordered, substrate-graphene interaction when it is grown on SiC, in this case when graphene is forced to grow over patterned SiC steps. This scalable bottomup approach allows us to produce a semiconducting graphene strip whose width is precisely defined within a few graphene lattice constants, a level of precision entirely outside modern lithographic limits. The architecture demonstrated in this work is so robust that variations in the average electronic band structure of thousands of these patterned ribbons have little variation over length scales tens of microns long. The semiconducting graphene has a topologically defined few nanometer wide region with an energy gap greater than 0.5 eV in an otherwise continuous metallic graphene sheet. This work demonstrates how the graphene-substrate interaction can be used as a powerful tool to scalably modify graphene's electronic structure and opens a new direction in graphene electronics research.Patterning a flat graphene sheet to alter its electronic structure was envisaged to be the foundation of graphene electronics. 1 The early focus was to open a finite-size gap in lithographically patterned nanoribbons, a necessary step for digital electronics. 1-5 While early transport measurements supported this possibility, 6 it soon became apparent that these "transport gaps" originated from a series of mismatched-level quantum dots caused by the inability of current lithographically to produce sufficiently narrow, well ordered, and crystallography define graphene edges. 7-10 A working solution to the graphene "gap problem" has yet to be formulated, let alone demonstrated. We show that in fact such a solution exists, not by patterning graphene, but instead by controlling the graphene-substrate geometry.We have been able to construct a unique, reproducible, and scalable semiconducting graphene ribbon with a gap larger than 0.5 eV. Using pre-patterned SiC trenches to force graphene to bend between a high symmetry (0001) face to a low symmetry (112n) facet, we produce a narrow curved graphene bend with localized strain. This "topologically-defined" ribbon is a wide-gap graphene semiconductor strip a few lattice constants wide that extends hundreds of microns long. The strip is connected seamlessly to metallic graphene sheets on both of its sides. One metallic sheet is n-doped and the other pdoped. From this simple morphology, we have not only produced a gap suitable for room temperature...