Crystallized silicon carbide (SiC) wafers are widely used in the field of integrated circuits as well as essential in the epitaxial growth of graphene and are one of the promising materials for applications in electronics at future high capacity. The surface quality of the required ultra-fine crystalline silicon wafer is the most essential factor in achieving graphene's desired electronic properties. Aiming to produce superfine surface quality SiC wafers, in this study, a new algorithm is developed to solve optimization problems with many nonlinear factors in ultra-precision machining by magnetic liquid mixture. The presented algorithm is a collective global search inspired by artificial intelligence based on the coordination of nonlinear systems occurring in machining processes. A new algorithm based on the optimization collaborative of multiple nonlinear systems (OCMNO) with the same flexibility and high convergence was established in optimizing surface quality when polishing the SiC wafers. To show the effectiveness of the proposed OCMNO algorithm, the benchmark functions were analyzed together with the established SiC wafers polishing optimization process. To give the best-machined surface quality, polishing experiments were set to find the optimal technological parameters based on a new algorithm and straight electromagnetic yoke polishing method. From the analysis and experimental results when polishing SiC wafers in an electromagnetic yoke field when using a magnetic compound fluid (MCF) with technological parameters according to the OCMNO algorithm for ultra-smooth surface quality with Ra=2.306 nm. The study aims to provide an excellent reference value in optimizing surface polishing SiC wafers, semiconductor materials, and optical devices