Do Virus Scanners Slow Down Your System?

[blueomni]

I must agree with some of the comments posted stating that a scan during game play can really bring the FPS down a lot. Too much disk IO.

This of course can be taken care of by NOT scheduling the system scan to run during your play time!!! duhh...

Now seriously. With a midrange dual core CPU at home and 4GB RAM I don't have any problems running MSE, maybe more bootup time and shutdown time required.

The main problem is really at the office. I use an IBM (Lenovo) laptop with a dual core and 3GB of ram. Our corporate AV is Sophos. Don't even try to compile anything with Eclipse, Sophos will scan every single file the compiler is trying to change and the process can take minutes! (Ok, you can say that corporate policies are not good...)

Bottom line, AV configuration is very important (scheduled scans, corporate policies ...) but also the PC capacities.
An article based on tests made on a high-end machine (at least for many of us) is not realistic.

 

[andregoujon]

ESET NOD32 please.

 

[dacian_herbei]

interesting article. I would not have thought to be true.
My personal experience with norton antivirus is quite the opposite.
What is missing in this review but maybe I missed it is Microsoft security essentials.
I and my family have this for at least one year and I'm quite satisfy with it.
It would have been nice to know how it behaves.

 

[Colossus]

As this entire article has pointed out, AV programs barely affect performance other then a second or 2 added to boot up and application startup. So to all of you saying you don't run any sort of AV software, do you really think it's better to be sorry rather than safe?

As several of you have said, IT savvy people who know not to click on the scary pop-up telling you to download this software now or DIE, don't tend to have virus problems. I don't think I've ever had any viruses and I don't see the need for entire, expensive suites such as Norton or McAfee, but I still have Windows Security Essentials running in the background, poised ready just in case the proverbial pooh-pooh hits the fan.

And as blueomni so cunning said, why schedule your scan to start when you're gaming? *tongue in bottom lip* UURGH! I don't tend to schedule at all, just scan every now and then when I think to myself "I haven't scanned in a while".

 

[schwizer]

Agreed i think Microsoft Security Essentials should be included. It's free, it's light and it works. It used to be Microsoft's Live OneCare if i'm not mistaken before they made it free. I've installed it on 5-6 different computers, replacing norton, AVG and others and it usually finds a few things the others did not.

 

[kikireeki]

NO Bitdefender! I'm really disappointed!
And if you want to see the real effect of an antivirus, try to extract a file with/without AV and eyewitness the startling difference!

 

[robertking82881]

try to run a full scan and run a game at same time on a single core. you it will be unplayable

 

[robertking82881]

try out julinux if you want to keep the windows look and convert easy . or try linux mint 10. if you cant do that then try justlinux or other ways run linux in vm ware and do all web surfing in linux. and scan stuff in linux . or you can duel boot but im to lazy for that lol

 

[robertking82881]

simple fix always log in as gust on your computer and log on to linux to scan computer befor switching to loging in as admin. you can be infected but noting can harm you logd in as guest

 

[C00lIT]

As an I.T. I will nitpick a little.
Great article and I agree, any fair dual core or single core manage quite well with almost all antiviruses.

But try comparing them on your NETBOOK using an Atom... You will see quite a difference.

You also left out the new Free Microsoft Security Essentials... and honestly I think it's the best Antivirus.

Also the best antivirus is the user's logic because most viruses and spyware... will still pass right by Avast, Norton and so on.

 

[davidja]

A great read. However, I think you should rerun this test on a system with less power. I am a field engineer and I find anti virus software really slows down the average system. Why don't you rerun you tested on a old P4 system with 1024MB of RAM. I would find those results interesting...

 

[avatar_raq]

Please add Microsoft security essentials to the mix..It's becoming more and more popular and it's free.
As a personal experience a laptop with a core 2 duo had a terrible lag with Norton internet (trial edition). This was resolved after uninstalling the damn thing and installing MSE instead.

 

[Guest]

How to take a simple question and make it a very complex one. Stop trying o show us how well you can make graphs and just cut to the chase. The answer is yes, any AV product that is running in the background uses some RAM and some CPU time. Get AVG 9 or the Kaspersky and quit worrying. You will be fine.

 

[jfby]

Where are the benchmarks for the computers infected with viruses? My idea of a 'study' is to have a control 'healthy' setup, setups with the different protections, and a controlled 'unhealthy' setup... seriously though it's nice to see that modern rigs really aren't slowed down noticeably.

 

[youssef 2010]

This is an excellent article as this question has been bothering me for a while but sadly, they didn't include Bit Defender in their antivirus list.
BTW,there's a typo in the test setting page as clearly you're using a Phenom II not an Athlon II

 

[jedi940]

I'm surprised nobody mentioned Zone Alarm. I have had good luck with it for years now. I would like to see results for that too.

 

[compton]

Having spent some time refreshing some XP/Pentium 4 systems, I'm not sure how to take these results. An older version of McAfee or Norton had brought the respective performace of these machines to a crawl; I though something was wrong. I certainly believe that Vista and 7 contribute to the lack of a performance hit, as well as optimizations in the security suites themselves. This leads to a problem. XP machines need these apps more, but don't usually have the hardware (or the newer versions) to run effectively. Even McAfee enterprise has a deleterious effect on XP installations, even on much newer harware. Still, this is another thoughful article, the kind of which more are desperately needed.

Keep it coming

 

[deputc26]

No MSE or Avast! ??? Useless, I thought was going to be great but you've omitted a lot of relevant software.

 

[cleeve]

No MSE or Avast! ??? Useless, I thought was going to be great but you've omitted a lot of relevant software.

Sorry, we would have liked to use everything but we had to draw the line somewhere. We used the most common AV software according to the numbers we could find, while it isn't perfect I think the cross section is wide enough that conclusions can be drawn.

 

[cleeve]

I must agree with some of the comments posted stating that a scan during game play can really bring the FPS down a lot. Too much disk IO.

Absolutely! A scan can significantly slow down even a quad core system. That's kind of a given and we weren't focused on that at all, but like the article says it's something we hope to explore in a follow-up article.

 

[cleeve]

I would have liked to have seen a seen a less high spec system or two thrown into the mix, something along the lines of a dual or single core...

I didn't make it clear in the specifications, but the majority of tests were run using DUAL-CORES.

I have modified the spec to clarify this. In addition, I encourage you to check out the single, dual, and quad-core performance comparison on page 7.

 

[noblerabbit]

Panda free Cloud Antivirus 2010, is highly recommended by me. I have tried many others, but this one is where I am staying at. recently updated. Cloud based, no slowdown, except perhaps a minor startup lag, even with SSD. But no more than 5 seconds.

 

[Alvin Smith]

WHO PAID YOU TO WRITE THIS?

I rely upon whatever free MS security center and update (removal tools) that are available, for free ... ONLY.

Security software is a SCAM and there is no way to tell if these very companies are not (will not) promulgate further exploits to ensure their own profits (kinda like drug companies and cancer research).

The ONUS ... The burden ... of sw security SHOULD BE on the OS vendors, the IP providers and upon Internet commerce, as well as upon governments ... NOT THE END USERS.

After all ... if governments (like the Chinese) and criminals (Internet commerce and bank/ident fraud) have large investments in hacking and botting large numbers of us, across the globe ... HOW CAN WE (and, why should we), as individuals, hope to win that war ??

Sure ... we will end up paying more taxes ... more for an OS and more for an internet transact but, the cost would be democratically distributed and the CONFLICT OF INTEREST would be significantly mitigated.

If the governments want their citizens to have a safe and secure internet ... If MS wants to sell me the next OS .... If Newegg/Amazon want my business ... then it is THEIR PROBLEM!!! ... NOT MINE!!

Put the BURDEN on those who are running the show and making all the money !

Security sw is for suckers and delerious dupes! If the big boyz want to use the web to get our business, then THEY are going to have to make the net safe and secure, for commerce and discourse. NOT MY JOB ... NOT MY PROB !!

All the fear and anxiety and manipulative marketing would disappear, at the very least and everyone, everywhere, will end up paying EXACTLY THEIR SHARE.

Give this some thot. If the web can't KEEP IT'S SELF SAFE, then, let's just not go there, until it is!

MS + US Cyber Command + IPs + Wal-Mart should be footing the bill ... LET *THEM* fit these costs into their own *COMPETITIVE* budgets and business plans.

SUCKERS!

 

[dikrek]

I think it really depends on whether the AV program checks the files for the disk benchmarks used in the test. If there's no difference, it simply means AV is NOT checking the files. It also depends on how many files are touched.

Using postmark, with extremely intensive settings (i.e. create thousands of files), the effects are pretty dramatic.

See benchmarks here: http://bit.ly/gnbnWN

 

[bildo123]

So they use the MSE logo at the end, but don't include it in the tests? Possibly one of the better, if not the best free AV solution out there. I myself have KIS 2011, which works well, but from checking the history on the other two computers it's installed on, MSE will do just fine after expiration.