As technology advances, Time-Dependent Dielectric Breakdown (TDDB) has become one of the major reliability threats for Copper/low-k interconnects. This article presents a novel approach, techniques, and flow for the physics-based chipscale assessment of backend low-k TDDB. In our work, the breakdown development is considered as the complementary combination of electric current path generation by means of diffusing metal ions and field-based hoping conductivity of the current carriers. It replaces the widely accepted acrosslayout electrostatic field based TDDB assessment. As a result, the model generated time-to-failure (TTF) is governed by kinetics of the electric current path generation, which is controlled by a time-dependent minimum metal ion concentration in the inter-metal dielectrics (IMD) gap-fill. Finite element analysis (FEA)-based simulations are used for populating the set of lookup tables, which provide a time to breakdown for any interconnect pattern with given geometries and voltages. A pattern-matching technique is used for extracting from the layout all patterns belonging to different classes of pattern shapes with different geometries, locations and electric loads. Experimental results obtained on a test chip show that upon the calibration the proposed flow provides a capability to evaluate chip-scale low-k TDDB reliability based on the calculated TTF and detect most leaking shapes in the layout.