5.11 PipelineModel

Use the PipelineModel plug-in to estimate the performance of workloads within a Fast Models environment.

The plug-in models first-order effects of microarchitecture components on the overall Cycles Per Instruction (CPI) value. Examples of such effects are data and structural hazards due to instruction dependencies.

The PipelineModel is implemented as a Fast Models plug-in. It processes instruction traces and injects cycle latencies into the simulation. The plug-in is non-intrusive to the functional accuracy of the simulation.

Figure 5-1 PipelineModel plug-in overview
To view this graphic, your browser must support the SVG format. Either install a browser with native support, or install an appropriate plugin such as Adobe SVG Viewer.

Fast Models provides the following prebuilt example PipelineModel plugins:


An approximation of the performance characteristics of the in-order, dual issue properties of the Cortex®‑A53 processor. The model provides estimated performance characteristics of a given compute-bound workload in terms of Cycles Per Instruction.


The model should not be used as a reference for hardware performance as it has limitations, such as the absence of a cache model.
An implementation of a single-issue 4-stage pipelined processor that illustrates the basic components of a CPU. It demonstrates how components such as Fetch, Decode, and Execute can be implemented as a Fast Models plug-in.

The source code is provided in %PVLIB_HOME%\plugins\source\PipelineModel\Cores\InOrder\. It contains a README and a makefile for building the example.


The model does not represent any Arm® core and is intended only as a guide for developing more advanced PipelineModels using Fast Models.

The PipelineModel is an aspect of Timing Annotation. For more details, see Timing Annotation in the Fast Models User Guide.

This section contains the following subsections:
Non-ConfidentialPDF file icon PDF version100964_1110_00_en
Copyright © 2014–2020 Arm Limited or its affiliates. All rights reserved.