Examination of 18 antivirus programs shows Microsoft Defender has the highest system load

[midian182]

Why it matters: A lot of people opt for Microsoft Defender as their default antivirus software, given that it comes as part of the Windows OS. But according to a new assessment of several different security packages, Microsoft's product uses up more system resources than other AV programs.

AV-TEST, an independent organization that evaluates and rates antivirus and security suite software, tested 18 antivirus packages for Windows 10. The examination involved testing the programs against around 12,000 malware samples mixed into 1.5 million files to see which ones could differentiate between something malicious and something innocent. It also checked for ransomware, though AV-TEST notes that not even a positive detection can prevent an attack from progressing.

The 18 security packages put through their paces were: AhnLab, Avast, AVG, Avira, Bitdefender, ESET, F-Secure, G DATA, K7 Computing, Kaspersky, Malwarebytes, McAfee, Microsoft Defender, Microworld, Norton, PC Matic, Protected.net and Trend Micro.

AV-TEST awarded a total of 6 points for each of the three test categories: protection, performance, and usability.

The good news for those who use it is that Microsoft Defender earned perfect scores of 6 in the protection and usability categories, but the 5/6 it was rated for in performance was the worst of all the programs tested.

"The highest system load in the test was generated by Windows Defender Antivirus for consumers. As the system load is considerably higher than that of the other products, Defender lost an entire point, thus ending up at 5 out of 6 points," wrote AV-TEST.

It wasn't explained in detail how much load Defender had on systems and just how badly it performed against the other antivirus programs.

Elsewhere, 14 of the 18 packages were awarded Top Product ratings for scoring 17.5 or higher. Microsoft Defender, ESET, Microworld, and PC all scored 17, but AV-TEST emphasized that these still offer very reliable levels of security, they just made a few minor errors.

The worst category score of all, a 3.0, was given to PC Matic in the useability section, earning it the lowest score overall (15). AV-TEST found that PC Matic identified well over two dozen false positives and blocked applications.

Six products managed to attain a perfect score of 18: Avast, Avira, Bitdefender, G DATA, Kaspersky, and Trend Micro.

Permalink to story.

https://www.techspot.com/news/98634-examination-18-antivirus-programs-shows-microsoft-defender-has.html

 

[DSirius]

Interesting results.
Microsoft Windows defender has only one advantage. It is the cheapest, you get it when you buy Windows License which costs 7-10 $ or Euro.

 

[Aus spot]

So it uses a few more resources. Not an issue if you have a few more core spare.

But really, next to all of them found all the viruses? How good is this test.

And false positives I know defender gives. So I don’t trust this report.

 

[Julnor]

Firewall is usually more important.

 

[ScottSoapbox]

Antivirus has been good for a while.

Which is why most attacks are aimed at the user rather than the computer.

 

[Nobina]

I'd like to see some benchmarks to see whether switching to a less resource hungry AV makes a difference. I doubt that it does, but I got no evidence.

I gave up on using 3rd party AV software 15 years ago, never had any issues like I used to back in Windows XP era. It seems that Windows Defender does basic security just fine and I don't notice it exists for the most part.

 

[Geralt]

With Kaspersky I am covered. Good!

 

[amoeba00]

Great. Just what I needed today - another rabbit hole for me to determine if Defender is good enough or if I should switch to another one..... and only to realize that same company, Gen Digital, owns Avira, AVG, and Avast (among others) - and that it used to be called Norton/Symantec.

 

[zulu53]

Another SaaS model bites the dust. Once the original AV program is created (most of these were created last century) they only need updates for "new" virus signatures that cannot be detected by the existing build in "virus behaviour" AI. How much is it really worth to pay for the very few "new" virus signatures? The answer is Windows Defender - it is worth $0.01. Purchasing virus protection software is indicative of a very poor risk reward model of the purchaser.

 

[antiproduct]

My university switched to Trellix (FireEye), I'm pretty sure that's much much worse.

 

[Hodor]

AI viruses are laughing at these antiviruses.

 

[Belgeek]

That might be the reason they don't filter emails on outlook.com anymore. It's too much load so it's better to let everything go in.

 

[WhiteLeaff]

AVG is good but drains the laptop battery

 

[RandomWAN]

How about some sort of reference/context to actually measure what the greater resource usage adds up to.

If it's a fractional difference during an active deep/full scan, whatever. In reality, I'm not going to be doing anything intensive on my machine during a deep/full scan anyway. Unless that adds up to it taking significantly longer to complete, I don't care.

If that's background resources, again, how much? Maybe it's 10% more than the others, but is that total amount going to be noticeable?

The only metric on these results that matters is the detection rate/success, the rest is all subjective noise without a system reference.

 

[texasrattler]

AVG is good but drains the laptop battery

Can't be all that good if it's killing the battery. It would make it useless for a person on the go.

 

[texasrattler]

So this article is rather useless and pointless without numbers being presented. What is the percentage between each AV? How much resouce/cpu is being used?
Does it affect any workload like gaming?

Any of this would be important info. Simply stating there is no info/data on how much resource is being used is like me turning in homework that's incomplete and expecting a passing grade. Not how it works. Data is key to articles like these, without it, it's a F.

While this article may not be TechSpot original work, either don't post these kinds until more data is given or if this is a original TS article, how did this pass getting posted without any data being involved?

 

[Eldritch]

While this article may not be TechSpot original work, either don't post these kinds until more data is given or if this is a original TS article, how did this pass getting posted without any data being involved?

Agreed. @Techspot may do a test of AV performance on benchmarks and games. That will be an interesting read.

 

[tristanj]

Tested on Windows 10? Okay... But what about testing on Windows 11 too for comparison purposes? Also, just a reminder to all those on 10, end of life is October 14, 2025. Time to move everything legacy that relies on Windows 10 is right now (if you haven't already).

 

[3volv3d]

Antivirus has been good for a while.

Which is why most attacks are aimed at the user rather than the computer.

Someone from Norton's said once that AV is pointless and takes a nice monthly yearly sub. Money for nothing.
It let's you know if you have a common threat.
If you scan your PC and it says 0. You feel comfortable.
While some guy changes the virus to just pass thru signatures.
Like you said user attacks.
Things will be better when online gaming goes offline.

 

[Alfatawi Mendel]

That might be the reason they don't filter emails on outlook.com anymore. It's too much load so it's better to let everything go in.

That must be why my hotmail account has been swamped with spam recently. I've had to migrate all of my stuff over to gmail. Bummer.

 

[Alfatawi Mendel]

Tested on Windows 10? Okay... But what about testing on Windows 11 too for comparison purposes? Also, just a reminder to all those on 10, end of life is October 14, 2025. Time to move everything legacy that relies on Windows 10 is right now (if you haven't already).

I'm in the process of moving to Linux. Fed up with being a Guinea Pig for Microsoft's broken updates, which benefit no-one but Microsoft.

 

[waclark]

"The highest system load in the test was generated by Windows Defender Antivirus for consumers. As the system load is considerably higher than that of the other products, Defender lost an entire point, thus ending up at 5 out of 6 points," wrote AV-TEST.

What is "considerably" higher? It used 5% CPU instead of 3%? The question is, how much does it impact overall performance? Is that impact during scans or is it more resource intensive when sitting in an "idle" state? I've used MS Defender for some time now and have never found it to be a problem. I've used other AV/Malware tools which seemed to have more impact on performance (looking at you Norton) but maybe that has all changed.

 

[waclark]

That might be the reason they don't filter emails on outlook.com anymore. It's too much load so it's better to let everything go in.

I have the opposite problem with Outlook. I get emails being sent to Junk/Spam that are from trusted users. Even when I select, "Trust all emails from ...." some still go into the Junk folder. Sometimes, some email from a specific user goes to junk but other emails, from the same user don't. It's crazy.

 

[Srksi]

Interesting results.
Microsoft Windows defender has only one advantage. It is the cheapest, you get it when you buy Windows License which costs 7-10 $ or Euro.

No, you get it for free even when you dont activate windows and you must actually pay to turn it off (*buy another antivirus) Its biggest malware ever

 

[trents]

The greater impact on system load of Microsoft Defender was not fleshed out with regard to real world impact on performance in this test. I doubt it's noticeable in real world computing, especially given the multi-core CPUs running today's systems. The fact that Microsoft Defender offers protection as good as the commercial options out there and is free seals the deal for me. And it doesn't nag you with "upgrade to premium protection" like other free products. Microsoft Defender will remain my AV of choice with Malwarebytes free as my second opinion AV checker product.