Skip to content

README.md

Latest commit

 

History

History
42 lines (29 loc) · 2.21 KB

README.md

File metadata and controls

42 lines (29 loc) · 2.21 KB

Plots

This directory contains both the scripts and data that we used to generate the plots and tables of our paper. The data needed to generate a plot can be found in the plot script's respective directory.

Reference Description Script
Fig. 6 Zooming out plot_trr_sampling_hist.py
Fig. 7 Zooming in: lightly sampled intervals plot_trr_sampling_hist_zoomedin.py
Fig. 8 Zooming in: ACT-based analysis plot_zoomin_exp.py
Fig. 11 Activation rate vs. hammering duration plot_trefi_time.py
Fig. 12 Required #tREFIs in sync for pattern P128 plot_hc_analysis_skh_128.py
Fig. 13 Comparison of refresh synchronization methods plot_prob_skh128_unified.py

Regenerating all plots

We provide a single script plot-all.sh that regenerates all the plots of our paper and saves them into the output/ directory.

Requirements

  • Python 3 (recommended: 3.8+)
  • Bash shell
  • The Python packages listed in requirements.txt
  • The script will automatically create and use a virtual environment in the plots/ directory
  • All plot scripts expect their data files to be present in the respective subdirectories

Running

Regenerate all plots by running:

chmod +x plot_all.sh
./plot_all.sh

Tables

Reference Description Data
Table 4 DDR5 RDIMMs: Pattern coverage and effectiveness fpga-results/
Table 5 Evaluation of DDR5 UDIMMs pc-results/