Looking under the hood of the IBM Blue Gene/Q network

Citations Over TimeTop 10% of 2012 papers

Abstract

This paper explores the performance and optimization of the IBM Blue Gene/Q (BG/Q) five dimensional torus network on up to 16K nodes. The BG/Q hardware supports multiple dynamic routing algorithms and different traffic patterns may require different algorithms to achieve best performance. Between 85% to 95% of peak network performance is achieved for all-to-all traffic, while over 85% of peak is obtained for challenging bisection pairings. A new software-controlled algorithm is developed for bisection traffic that selects which hardware algorithm to employ and achieves better performance than any individual hardware algorithm. The benefit of dynamic routing is shown for a highly non-uniform "transpose" traffic pattern. To evaluate memory and network performance, the HPCC Random Access benchmark was tuned for BG/Q and achieved 858 Giga Updates per Second (GUPS) on 16K nodes. To further accelerate message processing, the message libraries on BG/Q enable the offloading of messaging overhead onto dedicated communication threads. Several applications, including Algebraic Multigrid (AMG), exhibit from 3 to 20% gain using communication threads.

Related Papers

• Research and implementation of matrix transpose for real-time SAR imaging system(2011)
• → IKONIC IMAGERY IN THE SEVERELY SUBNORMAL(1972)3 cited
• Study on Matrix Transpose Algorithms(2011)
• → Adaptive Modified Transpose Jacobian Control of a Two-fingered Robotic Hand(2021)
• → Comparison between TRANSPOSE and MTF Algorithm with Closure Property(2023)