Application Sensitivity to Link and Injection Bandwidth on a Cray XT4 System
نویسندگان
چکیده
This paper describes our efforts to characterize application sensitivity to link and injection bandwidth on a Cray XT4 system. Link bandwidth is controlled by modifying the number of rails activated per network link. Injection bandwidth is controlled by modifying the speed of the HyperTransport connection between the Opteron and the SeaStar. A suite of micro-benchmarks and applications is evaluated at several different operating points. The experimental techniques developed by this work are expected to be useful for future architecture research.
منابع مشابه
Comparison of Cray XT 3 and XT 4 Scalability ∗
The Cray XT3 and XT4 have similar architectures, differing primarily in memory performance and in bandwidth between the node and interconnect. This paper evaluates and compares the scalability of the XT3 and XT4. Kernel benchmarks are used to verify and to quantify the performance differences between the systems. Application benchmarks are used to examine the impact of these differences on scal...
متن کاملAdvanced Communication Techniques for Gyrokinetic Fusion Applications on Ultra-Scale Platforms
In this paper we explore new parallel language constructs for the communication kernel of a real world magnetic fusion simulation code using the Partitioned Global Address Space (PGAS) model. The studied kernel is the particle shift phase of a tokamak simulation code in a toroidal geometry, which models the transit of charged particles between neighboring toroidal computational domains. We intr...
متن کاملUsing Processor Partitioning to Evaluate the Performance of MPI, OpenMP and Hybrid Parallel Applications on Dual- and Quad-core Cray XT4 Systems
Chip multiprocessors (CMP) are widely used for high performance computing. While this presents significant new opportunities, such as on-chip high inter-core bandwidth and low inter-core latency, it also presents new challenges in the form of inter-core resource conflict and contention. A challenge to be addressed is how well current parallel programming paradigms, such as MPI, OpenMP and hybri...
متن کاملThe Performance Effect of Multi-core on Scientific Applications
The historical trend of increasing single CPU performance has given way to roadmap of increasing core count. The challenge of effectively utilizing these multicore chips is just starting to be explored by vendors and application developers alike. In this study, we present some performance measurements of several complete scientific applications on single and dual core Cray XT3 and XT4 systems w...
متن کاملUnderstanding and Mitigating Multicore Performance Issues on the AMD Opteron Architecture
Over the past 15 years, microprocessor performance has doubled approximately every 18 months through increased clock rates and processing efficiency. In the past few years, clock frequency growth has stalled, and microprocessor manufacturers such as AMD have moved towards doubling the number of cores every 18 months in order to maintain historical growth rates in chip performance. This document...
متن کامل