2019arXiv (Cornell University)Open access

Modeling Shared Cache Performance of OpenMP Programs using Reuse Distance

Atanu Barai, Gopinath Chennupati, Nandakishore Santhi, Abdel‐Hameed A. Badawy, Stephan Eidenbenz

Open full text 1 citations

Abstract

Performance modeling of parallel applications on multicore computers remains a challenge in computational co-design due to the complex design of multicore processors including private and shared memory hierarchies. We present a Scalable Analytical Shared Memory Model to predict the performance of parallel applications that runs on a multicore computer and shares the same level of cache in the hierarchy. This model uses a computationally efficient, probabilistic method to predict the reuse distance profiles, where reuse distance is a hardware architecture-independent measure of the patterns of virtual memory accesses. It relies on a stochastic, static basic block-level analysis of reuse profiles measured from the memory traces of applications ran sequentially on small instances rather than using a multi-threaded trace. The results indicate that the hit-rate predictions on the shared cache are accurate.

Open-access reader

About this research paper

What this paper is about

Performance modeling of parallel applications on multicore computers remains a challenge in computational co-design due to the complex design of multicore processors including private and shared memory hierarchies. We present a Scalable Analytical Shared Memory Model to predict the performance of parallel applications that runs on a multicore computer and shares the same level of cache in the hierarchy. This model uses a computationally efficient, probabilistic method to predict the reuse distance profiles, where reuse distance is a hardware architecture-independent measure of the patterns of virtual memory accesses. It relies on a stochastic, static basic block-level analysis of reuse profiles measured from the memory traces of applications ran sequentially on small instances rather than using a multi-threaded trace. The results indicate that the hit-rate predictions on the shared cache are accurate.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Performance modeling of parallel applications on multicore computers remains a challenge in computational co-design due to the complex design of multicore processors including private and shared memory hierarchies. We present a Scalable Analytical Shared Memory Model to predict the performance of parallel applications that runs on a multicore computer and shares the same level of cache in the hierarchy. This model uses a computationally efficient, probabilistic method to predict the reuse distance profiles, where reuse distance is a hardware architecture-independent measure of the patterns of virtual memory accesses. It relies on a stochastic, static basic block-level analysis of reuse profiles measured from the memory traces of applications ran sequentially on small instances rather than using a multi-threaded trace. The results indicate that the hit-rate predictions on the shared cache are accurate.

Key concepts: Parallel computing, Computer science, Reuse, Cache, Computer architecture, Engineering, Waste management

Related papers

Back to paper searchBrowse research topicsOriginal source
Modeling Shared Cache Performance of OpenMP Programs using Reuse Distance — Research Paper | ScholarLens