Note on Optimal Trees for Parallel Hash Functions

TL;DR

This paper extends previous work on optimizing tree-based hash functions by introducing a simple algorithm that further reduces processor usage and computational work for more general tree topologies.

Contribution

It presents an algorithm to optimize tree topologies for parallel hash functions beyond the equal-depth leaf trees, improving resource efficiency.

Findings

Reduces number of processors needed for hashing

Decreases total computational work

Enhances efficiency of tree-based hash modes

Abstract

A recent work shows how we can optimize a tree based mode of operation for a rate 1 hash function. In particular, an algorithm and a theorem are presented for selecting a good tree topology in order to optimize both the running time and the number of processors at each step of the computation. Because this paper deals only with trees having their leaves at the same depth, the number of saved computing resources is perfectly optimal only for this category of trees. In this note, we address the more general case and describe a simple algorithm which, starting from such a tree topology, reworks it to further reduce the number of processors and the total amount of work done to hash a message.