{"id":69909,"date":"2014-07-29T10:59:29","date_gmt":"2014-07-29T08:59:29","guid":{"rendered":"http:\/\/onsoftware.en.softonic.com\/?p=69909"},"modified":"2025-07-02T00:34:46","modified_gmt":"2025-07-02T07:34:46","slug":"measuring-browser-speed-for-softonics-browser-comparison","status":"publish","type":"post","link":"https:\/\/cms-articles.softonic.io\/en\/measuring-browser-speed-for-softonics-browser-comparison\/","title":{"rendered":"Measuring browser speed for Softonic’s Browser Comparison"},"content":{"rendered":"

For Softonic’s Browser Comparison<\/a>, we compared Firefox<\/a>, Chrome<\/a>, Opera<\/a>, and Internet Explorer<\/a> using a methodology that’s easy enough for anyone to recreate themselves.<\/p>\n

Given most people’s heavy usage of the internet, browsers are increasingly one of the most important programs you have on your PC, and we wanted to test things which can be objectively measured<\/strong> but which are key when deciding which browser you should use. Most importantly, it’s considering whether a browser is both fast and light<\/strong>.<\/p>\n

We’ve designed a methodology to measure the performance that, in addition to being as objective as possible, aims to be easily understood <\/strong>by anyone who wants to recreate it themselves.<\/p>\n

The PC we used<\/h3>\n

The computer we used for all tests is a standard PC<\/strong> with 4 GB of memory. The operating system is 64-bit<\/a> Windows 7 Ultimate with all the updates, and it has Windows 7 because it’s the most popular OS.<\/p>\n

\"\"<\/p>\n

This is the PC we use for the tests; the hard drive is completely defragmented
\n<\/em><\/p>\n

During the test, we didn’t run other software (not even the antivirus) to avoid it interfering with the tests. All power-saving options were disabled<\/strong> so that the PC wouldn’t shut down due to inactivity.<\/p>\n

We restarted the PC <\/strong>before each test, and we left it for 30 seconds before logging in, which we timed to make sure that all tests were done on an even playing field.<\/p>\n

\"\"<\/p>\n

During the speed tests we don’t run any other program, not even an antivirus<\/em><\/p>\n

Loading pages from the Internet implies that speed will always change, so the pages we used for the speed test were loaded from a PC running Nginx<\/a>, a very fast and reliable web server.<\/p>\n

\"\"<\/p>\n

A local server avoids errors due to the reliability of the connection<\/em><\/p>\n

Browsers were always run in clean configuration with no history, cookies<\/a>, or extensions, to make sure that there was no code external to the browser running at the same time.<\/p>\n

Finally, the PC we used for the tests was connected to the Internet via a fiber optic connection. All tests are recorded with Adobe Premiere<\/a> and an HDMI <\/strong>capture.<\/p>\n

The tests we run for each browser<\/h3>\n

We chose to run only the tests that we considered to be\u00a0accurate <\/strong>and as close of a representation to its daily use as possible, which included looking at the speed when loading many tabs, and performance measurements that use graphs and common operations.<\/p>\n

\"\"<\/p>\n

Programs such as Peacekeeper test the page for common tasks.
\n<\/em><\/p>\n

The tests we carried out are:<\/p>\n

    \n
  1. Cold start-up with one tab <\/strong>on the home page of each browser. We measured the time it takes the browser to run and open Google after starting the computer.<\/li>\n
  2. Warm start-up with one tab<\/strong>. The same test was repeated half a minute after a cold start to see if the browser is faster after running it once.<\/li>\n
  3. Cold start-up with 11 tabs<\/strong>. 11 pages were loaded from the local web server and the clock stopped when the browser responded to scrolling using the mouse wheel.<\/li>\n
  4. Warm start-up with 11 tabs<\/strong>. Similar to the cold test with 11 tabs, we tested it 30 seconds after closing the browser from the cold start-up test.<\/li>\n
  5. Memory usage in megabytes<\/strong>. Memory usage was measured with 1 tab loaded, with 20 tabs loaded, and again with 1 tab loaded after closing the others, to verify how the memory recovers.<\/li>\n
  6. Speed tests with web technologies, <\/strong>which are used to create complex web apps. We chose to consider two independent tests which we believe are more reliable and closer to the real use:\n