The new method, called “Linear Complexity Multiplication” or “L-Mul, presents results close to floating point multiplication, but uses a simpler algorithm that requires fewer resources and, despite this, is capable of maintaining high precision and reliability necessary to enable artificial intelligence services.