Enhancing Parallel Game-Tree Searches by using Idle Resources of a High Performance Render Farm

Abstract

Due to the complex scheduling criteria, resource fragmentation often occurs in modern high performance computing systems. This paper addresses both of the policy and technical issues of using unstable resources formed by fragmentation. We propose that the parallel Monte-Carlo tree search (MCTS) can be implemented as lightweight single-core tasks to adapt to the unstable fractured resources. We develop a broker to collect idle resources in a high performance render farm and make use of them to execute MCTS tasks. The broker tends to assign promising tree nodes to reliable and responsive processors to confirm best-first search. Our work demonstrates a successful integration in which the MCTS computation gains significant resources without interfering with the render farm tasks.