mirror of
				https://github.com/openhwgroup/cvw
				synced 2025-02-11 06:05:49 +00:00 
			
		
		
		
	
		
			
				
	
	
		
			33 lines
		
	
	
		
			1.5 KiB
		
	
	
	
		
			Plaintext
		
	
	
	
	
	
			
		
		
	
	
			33 lines
		
	
	
		
			1.5 KiB
		
	
	
	
		
			Plaintext
		
	
	
	
	
	
| Wally PPA Study
 | |
| July 8, 2022
 | |
| Madeleine Masser-Frye
 | |
| mmasserfrye@hmc.edu
 | |
| ___________________
 | |
| Apologies for issues in this folder, code was written originally for individual use and documentation was compiled in haste. Please feel free to contact the author with questions.
 | |
| 
 | |
| -------------------
 | |
| ppaSynth.py
 | |
| 
 | |
| Run to synthesize datapath modules from src/ppa.
 | |
| To run a specific combination of widths, modules, techs, and freqs,
 | |
| modify those lists and use allCombos() to generate synthsToRun (comment out freqSweep).  
 | |
| To run a sweep of frequencies around the best delay found in existing syntheses (according to bestSynths.csv), modify the parameters and use freqSweep to generate synthsToRun.
 | |
| To remove synths to be run that already exist in /runs from synthsToRun, use filterRedundant().
 | |
| Syntheses run in parallel but you may encounter issues doing more than a dozen or so at once.
 | |
| -------------------
 | |
| ppaAnalyze.py
 | |
| 
 | |
| Run to plot results of PPA syntheses. See docstrings for individual function info.
 | |
| -------------------
 | |
| bestSynths.csv
 | |
| 
 | |
| Results of the synthesis for each combination of module, width, and tech with the best achievable delay.  Generated by csvOfBest() in ppaAnalyze.py
 | |
| -------------------
 | |
| ppaFitting.csv & ppaEquations.csv
 | |
| 
 | |
| Representations of the regression fit for each module and metric.  Generated in ppaAnalyze.py by makeCoefTable() and makeEqTable().
 | |
| -------------------
 | |
| ppaData.csv
 | |
| 
 | |
| Results from all synthesis runs.  Generated by synthsintocsv() and used by synthsfromcsv in ppaAnalyze.py.
 |