Optimising your PC’s performance can be a tricky and complex business, but the best place to start is by benchmarking its current speed, ideally with real-life tests such as the time it takes to launch some of your larger applications.
Once that’s done then it’s easy to measure how your system changes over time. Whether you’ve installed a new security suite which seems to have slowed you down, or a new defrag tool looks like it’s ramped up your speeds, then you won’t have to guess what’s happening; just run your tests again and you’ll see precisely what’s going on.
Of course it’s important that your benchmark is accurate, and that means hand-timing is out. It’s a much better idea to use an automated tool, like the free PassMark AppTimer, which can launch an application, then measure how long it takes to appear and poll for user input. Before closing the program and repeating the procedure, just as many times as you like, for improved accuracy.
To set the program up you must first specify the application you’d like to launch, and the log file where any data will be stored. AppTimer will use this to fire up your app, and start the clock.
Next, enter a word which appears in the program window’s caption bar. Normally the name of the application is enough: “Notepad”, “Word”, “Skype”, say.
And then tell AppTimer at what point you’d like it to stop the test clock. Typically you’d check “Input Idle”, “Window Name” and “Visible”, which means AppTimer will wait until a window has been created with a caption containing your specified word, it’s now visible, and waiting for user input. (But if that doesn’t work, clear “Input Idle” and try again.)
Once the time has been recorded and logged, then AppTimer will try to close the application for you. Check Alt+F4 to have the program simulate that keypress, for instance, or try the other two options as an alternative.
Finally, enter the number of times you’d like to repeat the launch test in the Executions box (we’d opt for at least 3, and ideally 5 to be sure you’re getting representative timings). And then specify the gap between each test in the Delay box, in milliseconds. By default this is only 1000 (1 second), but if it’s a bulky app, like Outlook, then you may want to extend this, giving the program time to close down properly: we’d use 10000 as a minimum.
And that’s it – you’re finished. Now close down any other programs you have running, click Run App, and watch as AppTimer launches, closes and relaunches your test application, just as many times as you’ve specified. Once it’s finished, open the log file to see the launch times. And if you repeat the test next week/ month/ year, or after installing an app, or running some speedup tool, you’ll be able to compare these times with the original set and look for any changes.
This process has its limitations, of course; measuring application launch times will primarily just tell you about hard drive and file system performance. And after the first test, it’s not even really doing that, as your application files will almost certainly be loaded from the Windows disk cache (so you may want to consider the first timing as the most important).
AppTimer has its usability issues, too. It really ought to save your chosen configuration options, for instance, so you can re-run the same tests without having to re-enter everything. And a button to display the log when the tests are complete would also be convenient.
Still, once you know what you’re doing then it only takes a moment to set up the program. And then AppTimer becomes a quick and easy way to measure application launch times. Its compact site (124KB) and portability are also plus points, and on balance we think the program merits a place in your benchmarking toolkit.