FEV can be broadly defined as comparing two models to determine whether they are equivalent. In general, it is customary to refer to the two models being compared as the SPEC, or specification modelSpecification model, and the IMP, or implementation modelImplementation model. Typically, the SPEC will be the more abstract model: it may be an RTL model, an unoptimized schematic netlist, a reference model, or a description in a high-level modeling language. The IMP will usually be an equally or more concrete model for comparison: a more refined model, a new or updated RTLRTL, or an optimized schematic netlist implementation.