norc 0.0.1

Noise Resilient Hardware Counters in HPC

NORC Analysis Components

NORC is a Python toolkit for measuring and analyzing the noise resilience of hardware counters in High-Performance Computing (HPC) systems.
It provides both command-line utilities and a GUI for automated performance measurements, statistical analysis, visualization, and ranking of metrics.

---

Installation

Clone the repository and install locally:

git clone https://github.com/tuda-parallel/NORC.git
cd NORC/analysis
pip install .

Analyzing Results

NORC offers two types of analysis tools:

• CLI Tools – norc_analyze, norc_plot, and norc_rank
• GUI Tool – norc_gui

NORC CLI

All CLI tools accept an experiment directory as input.
norc_plot and norc_rank must be run after norc_analyze has processed the experiment.

Example workflow:

norc_analyze /path/to/experiment     # Perform analysis
norc_plot /path/to/experiment        # Generate plots
norc_rank /path/to/experiment        # Rank metrics

NORC GUI

The GUI (norc_gui) requires no parameters. After launch:

1. Use the status bar to select the root directory of an experiment.
2. If the experiment has not been analyzed, NORC will process it automatically.

tdlr;

After installation, the following commands are available

pip install .
norc_gui # Launches the GUI
norc_analyze /path/to/experiment     # Analyze results
norc_plot /path/to/experiment        # Generate plots (requires analyze)
norc_rank /path/to/experiment        # Rank metrics (requires analyze)