I present a model of universal parallel computation called $Δ$-Nets, and a method to translate $λ$-terms into $Δ$-nets and back. Together, the model and the method constitute an algorithm for optimal parallel $λ$-reduction, solving the longstanding enigma with groundbreaking clarity. I show that the $λ$-calculus can be understood as a projection of $Δ$-Nets$-$one that severely restricts the structure of sharing, among other drawbacks. Unhindered by these restrictions, the $Δ$-Nets model opens the door to new parallel programming language implementations and computer architectures that are more efficient and performant than previously possible.