Guys, why not explain to jammydodger, if you can penetrate, that most software cannot divide the load between cores, and uses only one of the cores. The effective speed is, then, the speed of a single core.
It is unlikely that all cores will be run continuously, so multiplying the speed of a single core by the number of cores is silly. As is arguing over how it may be stated by the DxDiag or by a manufacturer.