In this work, we consider the possibility of building a magnonic co-processor for special task data processing. Its principle of operation is based on the natural property of an active ring circuit to self-adjust to the resonant frequency. The co-processor comprises a multi-path active ring circuit where the magnetic part is a mesh of magnonic waveguides. Each waveguide acts as a phase shifter and a frequency filter at the same time. Being connected to the external electric part, the system naturally searches for the path which matches the phase of the electric part. This property can be utilized for solving a variety of mathematical problems including prime factorization, bridges of the Konigsberg problem, traveling salesman, etc. We also present experimental data on the proof-of-the-concept experiment demonstrating the spin wave signal re-routing inside a magnonic matrix depending on the position of the electric phase shifter. The magnetic part is a 3 × 3 matrix of waveguides made of single-crystal yttrium iron garnet Y3Fe2(FeO4)3 films. The results demonstrate a prominent change in the output power at different ports depending on the position of the electric phase shifter. The described magnonic co-processor is robust, deterministic, and operates at room temperature. The ability to exploit the unique physical properties inherent in spin waves and classical wave superposition may be translated into a huge functional throughput that may exceed 1060 operations per meter squared per second for 50×50 magnetic mesh. Physical limits and constraints are also discussed.