Mavericks Browser Showdown: Safari 7 vs. Firefox 24 vs. Chrome 30

A new version of OS X has been released into the wild and, with it, a new version of Safari. I’ve had the luxury of playing with Mavericks during the beta and so I had been using Safari 7 off and on for a while. While I had already grown accustomed to a few of the new features (Shared Links and automatically stopping plug-ins to save power were the most noticeable), I wasn’t using it on my daily driver. So when it was released to the rest of the world, I finally got to see first-hand how much faster the computer felt due to the improvements in resource management.

I suffer from tab-creep in my web browsers. I have never implemented a proper system for taking sites I want to remember but don’t need right away and filing them away somewhere with an easy system for retrieving them. I used to bookmark everything and categorize it later. I’ve also tried dumping everything into Pocket, but it still mostly goes unused. So, as a result, I leave lots of tabs open in my browser until I get fed up and do something about the ones that are left open.

(Incidentally, if anyone out there has a suggestion on a service for filing, tagging, and searching sites I want to remember that is easily accessible from Mac and iOS devices, let me know.)

The first thing I noticed in Mavericks is how much faster Safari behaved with lots of tabs open. Many times I would have to quit Safari while I performed certain key tasks and then reopen it a few moments later when I was finished. This was even more aggravating due to the Internet at work running at glacial speeds during the majority of the day, as I often would have to buffer any videos I wanted to check out in the background in order to watch them later. Suddenly, however, I could leave Safari open and suffer no ill effects in other tasks. After reading up on how much more efficient Mavericks handles resources, I was excited, to say the least.

Throughout the last 5 years or so, several tech websites have performed “browser shootouts” on both Macs and PCs to break down the strengths and weaknesses of each web browser and declare one the victor over all. While Firefox and Chrome are updated semi-regularly, Safari only sees major improvements a few times a year. Generally, a new major release is refined over the course of the year, but adds very little in the way of new features, only bug fixes and optimization. One of the few times to truly see how the browsers compare to one another is to test them shortly after Safari’s major update, so I took it upon myself to do so.

What follows is a very simple test that is by no means exhaustive. But, it gives a good example of Mavericks’ impressive performance gains and what we can expect to see from Chrome and Firefox should they adopt the appropriate APIs to improve their resource management in Mavericks.

The Test

To compare the performance of each browser, I wanted something that was simple and fairly easy to replicate across each browser. So, I exported my bookmarks from Safari and imported them into Chrome and Firefox. Then, one at a time, I launched the browser and opened several bookmark folders into tabs. Once finished, I had 91 tabs open in the browser, four of which were YouTube, Vimeo, Hulu, and Netflix.

These final four tabs were where the real magic was going to happen. I started playing videos in each of the sites, and left Netflix as the front-most tab. Then, I jumped over to Activity Monitor to observe the results.

The first thing I noticed, in both Firefox and Chrome was how quickly the computer became taxed. The fan on the computer ramped up to full speed as it attempted to dispel the heat from my MacBook Pro. In Safari, however, the fan never sped up. I even performed the Safari test twice, once before the others and again at the end to ensure that the computer had been properly warmed up. The MBP kept its cool throughout Safari’s test both times.

I took screenshots of Activity Monitor and used a calculator to add up the percentage of CPU use and GBs consumed in RAM.

The Results

Here is what I discovered (all numbers are approximations):

  • Firefox used 262.6% CPU and 3.37 GB of RAM
  • Chrome performed a mite better using only 234.3% CPU and 3.17 GB of RAM
  • Safari blew them all away with only 85.7% CPU and 2.06 GB of RAM
  • Safari came in a very respectable first place with 120.6% CPU and 2.21 GB of RAM

(Shortly before publishing this, I realized that several instances of two processes that were owned by OS X but managed by Safari weren’t being counted in my initial results: com.apple.audio.ComponentHelper and com.apple.audio.SandboxHelper. Both appeared several times, but were only recognized as Safari processes when viewed hierarchically in Activity Monitor. I have included the correct results above.)

For Firefox, verification was easy. Since Firefox is the only one of the browsers tested that doesn’t separate each tab into its own set of processes, I could just add up the numbers for the Firefox app, and the two plugins being used: Flash and Silverlight.

Chrome and Safari were a bit more frustrating, as they break up tabs into multiple processes, rather than lumping them together. In fact, the first two times I ran this test for Safari, I missed some processes, as mentioned above. I might have missed some for Chrome, as well, but since the goal of this test was to see how Mavericks’ new APIs benefit resource management, I didn’t bother checking to see if Firefox actually managed to outperform Chrome. Once I had added up all the numbers for Chrome and Safari, it became evident that the new features in Mavericks were a huge benefit. It will be interesting to see if Chrome or Firefox see significant performance gains in future versions, as well.

Have you noticed any apps that perform drastically better in OS X Mavericks? Let me know in the comments.

Download this article as an e-book

This entry was posted in Apple, Random Acts of Technology. Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.
  • Subscribe to Blog via Email

    Enter your email address to subscribe to this blog and receive notifications of new posts by email.

    Join 282 other subscribers

  • Categories