Using ipywidgets and asyncio, it's easy to tap into the existing asynchronous event loop in Jupyter to make a cell that periodically monitors the output of a shell command (such as nvidia-smi) while you go about your work.

Asynchronously watch a shell command in a Jupyter Notebook.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import asyncio\n",
    "import subprocess\n",
    "import ipywidgets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def watch(cmd, interval=1, out=None):\n",
    "    '''Monitor the output of a shell command in a Jupyter Notebook asynchronously.\n",
    "    '''\n",
    "    out = out or ipywidgets.Output()\n",
    "\n",
    "    async def run():\n",
    "        while True:\n",
    "            stdout = subprocess.run(cmd.split(), capture_output=True).stdout.decode()\n",
    "            with out:\n",
    "                display(\n",
    "                    ipywidgets.HTML(\n",
    "                        f'<pre style=\"font-size: x-small; line-height: 1em\">{stdout}</pre>'\n",
    "                    )\n",
    "                )\n",
    "                out.clear_output(wait=True)  # clear upon receiving next output\n",
    "            await asyncio.sleep(interval)\n",
    "\n",
    "    task = asyncio.get_event_loop().create_task(run())\n",
    "    return out, task"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "806fe2fe65c84666af671c0cc59d251b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "out, task = watch('nvidia-smi')  # run task.cancel() to stop.\n",
    "out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# out, task = watch('df -H')\n",
    "# out, task = watch('ls -lt')\n",
    "# etc."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "@webio": {
   "lastCommId": null,
   "lastKernelId": null
  },
  "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}