On the design of a new dynamic credit-based end-to-end flow control mechanism for HPC clusters

Abstract

Adaptation of the credit-based flow-control mechanism to the EXTOLL interconnect.In-depth evaluation of the previous flow control including the use of piggybacking.Design of a lightweight dynamic credit-based flow-control mechanism for clusters.Thorough performance evaluation of the new dynamic flow control.Analysis of the memory footprint of both flow control versions. High Performance Computing usually leverages messaging libraries such as MPI, GASNet, or OpenSHMEM, among others, in order to exchange data among processes in large-scale clusters. Furthermore, these libraries make use of specialized low-level network layers in order to achieve as much performance as possible from hardware interconnects such as InfiniBand or 40Gb Ethernet, for example. EXTOLL is an emerging network targeted at high performance clusters.Specialized low-level network layers require some kind of flow control in order to prevent buffer overflows at the receiver side. In this paper we present a new end-to-end flow control mechanism that is able to dynamically adapt, at execution time, the buffer resources used by a process according to the communication pattern of the parallel application and the varying activity among communicating peers. The tests carried out on a 64-node 1024-core EXTOLL cluster show that our new dynamic flow control mechanism presents very low overhead with an extraordinarily high buffer efficiency, as overall buffer resources are reduced by 4i? with respect to the amount of buffers required by a static flow control protocol achieving similar low overhead levels.