AV-Comparatives Performance Test - April 2022

The comment by @moredhelfinland is clear to me. It basically says (in so many words) that: (a) AV-C's Performance test shows that MS Defender is a bit heavier on system resources than is the case for several other AVs, and (b) Windows integral FW & some of its 3rd party "front end" apps are not well protected.

That poster is a relatively new member and I have learned useful info from several of that member's posts. Obviously, English is not that member's native tongue. In formal English, the word "stupid" can apply only to sentient persons (see HERE), and computer apps are not sentient persons --- yet.

 

I doubt, that AV-Comparatives testing methodology is relevant for practical use.

My own experience, with hundreds of customer machines,
shows, that performance issues are not caused by MD, and neither the hardware,
but the history of the installed Windows.

Sadly fact:
AV hobby is boring, without MD bashing...

 

In #26, I posted:

Thus, I merely posted a comment based on AV-C's test-based scores. That comment is fact-based & on-topic, not "bashing."

AV-C's lab-based tests are statistically significant, based on a specific data set at a specific point in time. Hence those lab-based tests, while not omniscient, are far better than simplistic anecdotal comments.

I do not have an "AV hobby." Rather, I have a strong interest in keeping several computers (one of which is my own) up & running, free of infection, and sprightly in performance. So I find AV-C's various tests to be very useful. If someone believes they are useless tests, compared with anecdotal comments, I wonder why they visited this particular thread in the first place.

 

They are mostly useless. How can you base any test run on one machine and say that it applies to the other millions of computers out there with billions of hardware and software combinations? What applies to one computer does not apply to another. There are many posts on here and other forums that say WD does not slow there machine down, and other posts that say that it does, and other posts that say other AV's slow there machines down. It comes down to what WD and any other program does on YOUR computer, your results do not mean that they apply to any other machines. These tests, AV protection test etc, should be taken as entertainment value only as there are too many variables to say anything with certainty and anyone who has been a member here for any length of time should know that.

 

Again, I wonder why you're even visiting a thread about what you evidently consider is a worthless test.

AV-C's statistical base is sound. Its staff members are professionals. I have yet to see your credentials. In any event, any sampling test is always an indicator. Gathering universe-wide data is infeasible as to both costs & practicality. Thus, the alternative to considering tests based on statistically valid samples is to make choices blindly, based on anecdotal claims by those who say their unsupported assertions are more valid than professionally designed tests that fully disclose their structural parameters.

I have been a member here for a good while and yet I disagree with you. I wonder what's your need for the implied insult of long-time members who frequent threads about AV-C's test results? I also wonder why you feel the need to post these same sort of negatory comments in just about every AV-C test thread?

 

Well if we look at results MD is slowing down file copying and application installation processes but not other - launching apps, browsing websites...
From personal experience file copying is indeed greatly impacted by MD real-time protection. I never tested installing apps with or without protection so can't say about that.
So IMO regular daily usage will probably not show any system slowdown. But when I mange systems with MS enabled I always disable it's real-time protection during file related operations (making backups...). Increase in speed can be enormous.

 

I find it amazing that almost any and every AV thread can come back to bashing MD, but that's just me.

 

https://en.wikipedia.org/wiki/Opinion_poll

 

@digmor crusher et alia -- Sorry if I was unclear. My particular degree had a prerequisite for an advanced math course, even though my degree had very little to do with math. I think that prerequisite was put there in order to "thin the herd."

I chose probabilities & sampling theory for the advanced math. It was interesting but never earned me a penny in the career path that eventuated from my Navy assignments during a certain war. Most of the terms I used in post #30 are explained HERE, except that the linked article uses "population" for that factor which my post #30 calls "universe." Universe is the more formal term.

I had no intent to yank anyone's chain but felt it necessary to rebut misinformation. AV-C's tests are useful tools for choosing & checking security configurations. The reports covering each & every one of AV-C's manifold tests are all preceded by detailed definitions of methodologies, sampling base, equipment used, etc. That detailed information enables a knowledgeable interpretation of any given test's strength & weaknesses.

The participants in AV-C's tests pay for the privilege, partly for "bragging rights" (perhaps) but also because those tests help them to assess their product's weaknesses & strengths & thereby improve their products. Thus, the tests are not toys for entertainment. Rather they are very useful tools for informed proponents & users of security software.

 

the funny thing beside avc comparisons is that they never tell about security isses of antivirus itself.

eg avast and avg fixed now a security issue which is 10 years old. this flaw can lead to have malware access with system rights because direct access to the kernel, without user interaction. and such malware, i think about exploit kits, are tested regularly against all leading antivirus protections not to get caught. kaspersky has or had similar issues.

detection rates are overrated, ok, MBAM is really bad, but at least the leading vendors are around 99.x.

we have currently "560,000 new pieces of malware" each day.
https://dataprot.net/statistics/malware-statistics/
in 2020 it was "only" "360000 new malicious files were detected by Kaspersky per day"
https://www.kaspersky.com/about/pre...d-every-day-increases-by-52-to-360000-in-2020

if you are sitting on a saw you dont remember which tooth hit you.

antivirus suites need to be full featured because they need to replace windows defender and its firewall at all. defender is no longer a single "stupid" antivirus, it has any feature which others offers too.
but it is not advertised so most people do not know about such features.

using windows 7 and got sticked with it you probably never would get it, windows 8 was better, win10 pretty good, win11 best.

thoses daily (!) biased tests are nice, but nothing more. some users do not notice impacts of their used antivirus - in case of impacts you should remember threads like this and think again if the chosen product could be the right one.

 

They're testing on a system with Core i3 and 4GB of ram, it's intended to be a "low end" system. I never had a performance issue with Defender on my 8700K, or my wife's R5 1600X. Honestly, if you're running 4gb of ram, it's time to upgrade. Something. Somewhere.

 

on a dual core cpu you sure will have a slow down, but no longer quad core or above.
4gb ram is quite enough to start, but for working 8gb or more is recommended. for modern games 32gb and more, graphics need 16gb or more (32gb video ram = better).
but the longer defender is running here, the less the slow down. it takes time to have all files scanned, not only in background.

 

Thanks, no problem.

 

@digmor crusher -- Live long & prosper!

@Brummelchen -- very informative posts!!!

I think the member who used the word "stupid" is less fluent in English than you are, & he might not realize how harsh and demeaning it is to call something "stupid."

 

According to AV-test.org, MD is faster than average and does not slow down,
weither on slow , or fast machines.
Amen.

 

According to AV-test.org, MD is faster ALL ARE FASTER than average and does not slow down, weither on slow , or fast machines. Amen.

In "Performance", show me one with less than 6 =
https://www.av-test.org/en/antivirus/home-windows/
6 = Outstanding / 3= Satisfactory / 0 = Insufficient

 

AV-TEST results are BS. All AVs have a high performance rating. Trust who you want.

 

Last one laughing:

 

One can find the methodology used for the above test on pages 43-44

 

You are talking for apples (AV-test.org) and then you show oranges (mrg-effitas)...
Both are BS, two minus make one plus only in maths.............

 

I checked methodology and it doesn't really say how many files they used for file copy operation test (weak spot of MD). They all achieved times from 1,8 to 2,4 seconds. Which makes me think they've used 5-10 files to test copy speed. IMO that's not enough to test impact on file copy operations. They should try copying few 100s or 1000s of files and difference should be shown more clearly.

 

From an empirical standpoint we could never know the efficacy of any AV test. So all debates are ultimately speculation. I don't really care about any of it because a daily image is my first and last line of defense. Also, the differences are now all marginal between them so let's all calm down and stay civil.

 

AV Test is a joke. 18 products tested, 14 are "TOP PRODUCTS!", and of the ones that aren't: one is a scam product, another can't even design a webpage that fits the width of your screen, one is "one of the top AVs in Japan!", and the last one doesn't even load properly if you're using an ad blocker. GTFO with that crap.

BTW, let's make a fun game and guess which of the four are which.

 

+1