The feasibility of using metal optics or negative ε materials, with the aim of reducing the transversal extent of waveguided photonic fields to values much less than the vacuum wavelength, in order to achieve significantly higher densities of integration in integrated photonics circuits that is possible today is discussed. Relevant figures of merit are formulated to this end and used to achieve good performance of devices with today’s materials and to define required improvements in materials char- acteristics in terms of decreased scattering rates in the Drude model. The general conclusion is that some metal based circuits are feasible with today’s matals. Frequency selective metal devices will have Q values on the order of only 10~100, and significant improvements of scattering rates or lowering of the imaginary part of ε have to be achieved to implement narrowband devices. A photonic “Moore’s law” of integration densities is proposed and exemplified. The feasibility of using metal optics or negative ε materials, with the aim of reducing the transversal extent of waveguided photonic fields to values ​​much less than the vacuum wavelength, in order to achieve significantly higher densities of integration in integrated photonics circuits that is possible today is discussed. Relevant figures of merit are formulated to this end and used to achieve good performance of devices with today’s materials and to define required improvements in materials char- acteristics in terms of decreased scattering rates in the Drude model. The general conclusion is that the metal based circuits are feasible with today’s matals. Frequency selective metal devices will have Q values ​​on the order of only 10 ~ 100, and significant improvements of scattering rates or lowering of the imaginary part of ε have to be achieved to implement narrowband devices. A photonic “Moore’s law” of integration densities is proposed and under.