NVIDIA Nsight Systems Recipes

nsys_kernel_stats.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import sqlite3\n",
    "\n",
    "kernels = pd.read_sql_query(\"select StringIds.value as name, num, min as 'min (ns)', max as 'max (ns)', avg as 'avg (ns)', total as 'total (ns)' from cudaKernelStats INNER JOIN stringIds on stringIds.id == shortName order by total desc\", sqlite3.connect(\"report1.sqlite\"))\n",
    "kernels[\"min (s)\"] = kernels[\"min (ns)\"] / 1e9\n",
    "kernels[\"max (s)\"] = kernels[\"max (ns)\"] / 1e9\n",
    "kernels[\"avg (s)\"] = kernels[\"avg (ns)\"] / 1e9\n",
    "kernels[\"total (s)\"] = kernels[\"total (ns)\"] / 1e9\n",
    "kernels"
   ]
  }
 ],
 "metadata": {
  "hide_input": false,
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}

nsys_kernel_stats.py

import pandas as pd
import numpy as np
import sqlite3

kernels = pd.read_sql_query("select StringIds.value as name, num, min as 'min (ns)', max as 'max (ns)', avg as 'avg (ns)', total as 'total (ns)' from cudaKernelStats INNER JOIN stringIds on stringIds.id == shortName order by total desc", sqlite3.connect("report1.sqlite"))
kernels["min (s)"] = kernels["min (ns)"] / 1e9
kernels["max (s)"] = kernels["max (ns)"] / 1e9
kernels["avg (s)"] = kernels["avg (ns)"] / 1e9
kernels["total (s)"] = kernels["total (ns)"] / 1e9
print(kernels)