Do Security Solutions Slow Down Your PC? (Part 10 – BitDefender Internet Security 2012)

Our series, “Do Security Solutions Slow Down Your PC?” has been a huge hit. We’ve gotten great feedback from all around the world, both from consumers and even some software makers. Now that security companies are releasing their 2012 products, we decided to give this series a complete reboot. After all, it wouldn’t be unfair to pit a 2012 security suite against a 2011 version. Plus, some of the benchmark tools we used have been significantly updated over the past few weeks.

We’re going to start by not just outlining a new set of tools and rules, but by showing the test results for a security suite you’ve been clamoring for for months now: BitDefender Internet Security 2012. According to the AV-Test Institute in August 2011, this product is one of the best suites on the market. Next to Kaspersky, BitDefender was the only contender (out of 25!) to receive a perfect detection rate (100%).

So this begs the question: Do high detection capabilities compromise PC performance, or is BitDefender perfect in this area, as well?

A new test rig

I said goodbye to my trusty old MacBook Air (Core 2 Duo 1.86 GHz, 4 GB of RAM and GeForce 9400M) and replaced it with something a bit faster: a 2009 laptop that sports a Core 2 Duo 2.93 GHz, 4 GB of RAM, a 256 GB hard disk and a GeForce 9600M GT. This nearly three-year old rig should reflect the average performance of most mid-level PCs out there. It should also demonstrate the impact security solutions could have on systems that are more modern.

To get the most accurate results, I used AppTimer which automatically launches and closes applications to measure start-up times down to the millisecond.

We set AppTimer up to start programs (such as Internet Explorer 9, Outlook, and PhotoImpact) to launch five times with a delay of five seconds in between. We also evaluated the first launch (“Cold Start”) and subsequent launches (“Warm Starts”) loaded from memory and were thus, faster. Let’s see how BitDefender affected the new test PC and benchmarking system.

Timecode 00:00:00 – Scenario starts / Boot performance

To determine boot time, we first measured how many seconds it took from pressing the power button until the moment the Windows 7 desktop appeared. Then, we measured how long it took until Windows Live Messenger (an auto-start application) was visible on the screen, and used Windows Performance Toolkit to determine the exact time needed for all services and processes to finish booting.

In our previous tests, security suites delayed the boot time of Windows 7 quite a bit. So how did BitDefender fare?

Quite well actually! Before installing BitDefender, the desktop appeared after 37 seconds. Fortunately, the security suite added just five seconds on top of that which is barely noticeable. Windows Live Messenger, however, showed up 12 seconds later. Overall, boot time lagged by 6.5 seconds. Bear in mind that on slower systems, this result might be more drastic.

Timecode 00:10:00 – Resource check

Ten minutes after booting, we launched Windows Task Manager and checked on the resources BitDefender ate up.

The security suite increased total memory consumption by almost 300 MB which we found to be just a bit too much. This makes quite a difference for a system with a low amount of memory. We also thought that seven additional active processes were way too many yet suspect this is necessary to ensure full protection.

Timecode 00:11:30 – Browser start

We  switched from testing Google Chrome’s startup to Internet Explorer 9′s.

Thanks to AppTimer, we were able to determine both the “Cold Start” and the “Warm Start”.

AppTimer showed a 0.1 and 0.3 delay respectively which is not at all noticeable. Good job to BitDefender!

Timecode 00:12:30 – Launching Outlook

Outlook loads dozens of libraries, plug-ins, and one large file (the Exchange OST file in our test case) into memory.

Again, the results are barely noticeable. In contrast to the many other security solutions previously tested, the e-mail plug-in didn’t slow things down at all. Bravo!

Timecode 00:14:30 – Launching PhotoImpact X3

Our favorite photo editing tool, PhotoImpact, is heavier than Internet Explorer 9 and Outlook combined. Usually, it takes about 3 to 3.3 seconds to load—let’s see if and how BitDefender changed this.

Security solutions should not delay program launches, but BitDefender did. Although we’re only talking about a delay of 1.3 and 0.7 seconds respectively, on slower systems, this might double the regular amount of time needed.

Timecode 00:19:30 – Converting an audio clip from WMA into AAC with iTunes

iTunes’ audio conversion takes a heavy toll on the CPU. We wanted to know how much of an impact BitDefender had on pure processing performance when converting a ten-minute audio clip from Microsoft’s WMA format into AAC.

Luckily, there was no discernible effect. The conversion took 22 seconds to complete with and without BitDefender installed.

Timecode 00:35:00 – Large file extraction

Extracting ZIP files needs both CPU and hard disk power. To test extraction speeds, we compressed a folder with more than 1000 documents and TIF pictures into one 600 MB ZIP file. Then, we measured how long it took to unzip this file.

In each test run, we noticed a slight delay of about two seconds. This isn’t anything to write home about, though.

Timecode 00:47:00 – PCMark 7 results

PCMark 7 tests a variety of critical performance areas, such as file copying, text editing, file compression, etc. The new and improved version is more accurate and covers more realistic test scenarios than its predecessor (PCMark Vantage).

PCMark 7 showed a slight decrease in performance. Web browsing, in particular, took a bit of a hit (about a 10% performance loss), but is this really that significant?

Performance results: BitDefender Internet Security 2012

Overall, we were very pleased with BitDefender. It didn’t just get some of the highest grades over at AV-Test.org, its impact on performance was barely noticeable. We are very curious to see if the other contenders handle this new test setup equally well.