PRIMA is a novel algorithm for generating provably passive reduced-order N-port macromodels for RLC interconnect circuits. It extends the Block Arnoldi technique to ensure passivity, leading to more accurate results than existing methods. PRIMA uses congruence transformations to preserve passivity during reduction, ensuring that the resulting models are stable and passive. The algorithm efficiently computes reduced-order macromodels for generalized RLC interconnects, using path tracing for high efficiency. PRIMA is compared with other methods like MPVL and Block Arnoldi, showing comparable or superior accuracy. It is demonstrated that PRIMA produces stable poles and accurate results, even for complex circuits. The algorithm is integrated with RICE for efficient moment calculation and is shown to significantly improve simulation performance. Results show that PRIMA provides accurate high-frequency responses and is efficient for large circuits. The method is proven to preserve passivity, making it suitable for time-domain analyses. PRIMA can be extended to include heuristics like moment shifting and frequency shifting, though these are not strictly necessary. The algorithm is effective for a wide range of RLC interconnect problems, including mesh ground planes, transmission lines, and coupled noise scenarios. Overall, PRIMA offers a reliable and efficient approach for generating passive reduced-order macromodels for RLC circuits.PRIMA is a novel algorithm for generating provably passive reduced-order N-port macromodels for RLC interconnect circuits. It extends the Block Arnoldi technique to ensure passivity, leading to more accurate results than existing methods. PRIMA uses congruence transformations to preserve passivity during reduction, ensuring that the resulting models are stable and passive. The algorithm efficiently computes reduced-order macromodels for generalized RLC interconnects, using path tracing for high efficiency. PRIMA is compared with other methods like MPVL and Block Arnoldi, showing comparable or superior accuracy. It is demonstrated that PRIMA produces stable poles and accurate results, even for complex circuits. The algorithm is integrated with RICE for efficient moment calculation and is shown to significantly improve simulation performance. Results show that PRIMA provides accurate high-frequency responses and is efficient for large circuits. The method is proven to preserve passivity, making it suitable for time-domain analyses. PRIMA can be extended to include heuristics like moment shifting and frequency shifting, though these are not strictly necessary. The algorithm is effective for a wide range of RLC interconnect problems, including mesh ground planes, transmission lines, and coupled noise scenarios. Overall, PRIMA offers a reliable and efficient approach for generating passive reduced-order macromodels for RLC circuits.