Analytical Performance Estimation for Large-Scale Reconfigurable Dataflow Platforms
- 12 August 2021
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 14 (3), 1-21
- https://doi.org/10.1145/3452742
Abstract
Next-generation high-performance computing platforms will handle extreme data- and compute-intensive problems that are intractable with today’s technology. A promising path in achieving the next leap in high-performance computing is to embrace heterogeneity and specialised computing in the form of reconfigurable accelerators such as FPGAs, which have been shown to speed up compute-intensive tasks with reduced power consumption. However, assessing the feasibility of large-scale heterogeneous systems requires fast and accurate performance prediction. This article proposes Performance Estimation for Reconfigurable Kernels and Systems (PERKS), a novel performance estimation framework for reconfigurable dataflow platforms. PERKS makes use of an analytical model with machine and application parameters for predicting the performance of multi-accelerator systems and detecting their bottlenecks. Model calibration is automatic, making the model flexible and usable for different machine configurations and applications, including hypothetical ones. Our experimental results show that PERKS can predict the performance of current workloads on reconfigurable dataflow platforms with an accuracy above 91%. The results also illustrate how the modelling scales to large workloads, and how performance impact of architectural features can be estimated in seconds.Keywords
Funding Information
- EU H2020 Research and Innovation Programme (671653)
- UK EPSRC (EP/P010040/1, EP/N031768/1, and EP/L016796/1)
- JST/CREST program “Research and Development on Unified Environment of Accelerated Computing and Interconnection for Post-Petascale Era”
- JSPS KAKENHI (20K19770)
This publication has 31 references indexed in Scilit:
- COMPASSPublished by Association for Computing Machinery (ACM) ,2015
- ExaSAT: An exascale co-design tool for performance modelingThe International Journal of High Performance Computing Applications, 2015
- Performance Modeling for FPGAs: Extending the Roofline Model with High-Level Synthesis ToolsInternational Journal of Reconfigurable Computing, 2013
- What It'll Take to Go ExascaleScience, 2012
- GPURoofline: A Model for Guiding Performance Optimizations on GPUsLecture Notes in Computer Science, 2012
- Performance Analysis Framework for High-Level Language Applications in Reconfigurable ComputingACM Transactions on Reconfigurable Technology and Systems, 2010
- An analytical model for a GPU architecture with memory-level and thread-level parallelism awarenessACM SIGARCH Computer Architecture News, 2009
- RooflineCommunications of the ACM, 2009
- Basic Terminology, Notation and ResultsPublished by Springer Science and Business Media LLC ,2009
- Parameterized dataflow modeling for DSP systemsIEEE Transactions on Signal Processing, 2001