rename to power-frequency model

Author: benvanwerkhoven

CHANGELOG.md

@@ -6,6 +6,8 @@ This project adheres to [Semantic Versioning](http://semver.org/).
 
 ### Added
 - Support for using time_limit in simulation mode
+- Helper functions for energy tuning
+- Example to show ridge frequency and power-frequency model
 
 ### Changed
 - Changed what timings are stored in cache files

examples/cuda/going_green_performance_model.py

@@ -1,14 +1,14 @@
 #!/usr/bin/env python
 """
-This example demonstrates how to use the performance model presented in
+This example demonstrates how to use the power-frequency model presented in
 
   * Going green: optimizing GPUs for energy efficiency through model-steered auto-tuning
     R. Schoonhoven, B. Veenboer, B. van Werkhoven, K. J. Batenburg
     International Workshop on Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS) at Supercomputing (SC22) 2022
 
 to reduce the number of frequencies for GPU energy tuning.
 
-In particular, this example creates a plot with the modeled-performance vs
+In particular, this example creates a plot with the modeled power consumption vs
 frequency curve, highlighting the ridge frequency and the frequency range
 selected by the user.
 
@@ -59,7 +59,7 @@ def get_default_parser():
     parser = get_default_parser()
     args = parser.parse_args()
 
-    ridge_frequency, freqs, nvml_power, fitted_params, scaling = energy.create_performance_frequency_model(device=args.device,
+    ridge_frequency, freqs, nvml_power, fitted_params, scaling = energy.create_power_frequency_model(device=args.device,
                                                                                                n_samples=args.samples,
                                                                                                verbose=True,
                                                                                                nvidia_smi_fallback=args.nvidia_smi_fallback,

kernel_tuner/energy/energy.py

@@ -116,7 +116,7 @@ def estimated_power(X, clock_threshold, voltage_scale, clock_scale, power_max):
     return powers
 
 
-def fit_performance_frequency_model(freqs, nvml_power):
+def fit_power_frequency_model(freqs, nvml_power):
     """ Fit the performance frequency model based on frequency and power measurements """
 
     nvml_gr_clocks = np.array(freqs)
@@ -149,10 +149,10 @@ def fit_performance_frequency_model(freqs, nvml_power):
     return clock_threshold + clock_min, fit_parameters, scale_parameters
 
 
-def create_performance_frequency_model(device=0, n_samples=10, verbose=False, nvidia_smi_fallback=None, use_locked_clocks=False):
+def create_power_frequency_model(device=0, n_samples=10, verbose=False, nvidia_smi_fallback=None, use_locked_clocks=False):
     """ Calculate the most energy-efficient clock frequency of device
 
-    This function uses a performance model to fit the performance-frequency curve
+    This function uses a performance model to fit the power-frequency curve
     using a synthethic benchmarking kernel. The method has been described in:
 
      * Going green: optimizing GPUs for energy efficiency through model-steered auto-tuning
@@ -186,7 +186,7 @@ def create_performance_frequency_model(device=0, n_samples=10, verbose=False, nv
         print("Clock frequencies:", freqs.tolist())
         print("Power consumption:", nvml_power.tolist())
 
-    ridge_frequency, fitted_params, scaling = fit_performance_frequency_model(freqs, nvml_power)
+    ridge_frequency, fitted_params, scaling = fit_power_frequency_model(freqs, nvml_power)
 
     if verbose:
         print(f"Modelled most energy efficient frequency: {ridge_frequency} MHz")