AV-test.org issues latest figures

In-depth testing covers multiple factors.

Independent testing body AV-Test.org has released its latest set of results, with a large group of products tested against a number of criteria including proactive detection, spotting and removing active infections, and outbreak response times, as well as simple detection rates.

The results show how companies and their products fare against the latest range of samples arriving at AV-Test, with results of checking new arrivals used to determine the accuracy of heuristics and the efficacy of behavioural detection systems. Updates were also monitored over the test period to determine when companies added detection for new items not spotted using heuristics or generic detection. Detection and effective removal of active malware, including rootkits, is also measured, as is the impact on system performance.

As in AV-Comparatives' recent figures, multi-engine products such as AEC's Trustport, G DATA's AVK and the gateway scanning product WebWasher all performed very strongly in the pure detection test, with Avira's AntiVir also achieving very high scores in both malware and 'potentially unwanted' categories.

The multi-engine products showed their weakness when it came to scanning times and false positives however, and also fared poorly against rootkits, while Avira did well across the board, ranking 'good' or 'very good' in all categories. The only other product to achieve this feat was Sophos, with Symantec and Panda let down only by their response times to outbreaks, marked as merely 'Satisfactory', and McAfee also failing to excel in scanning speed.

The results of the tests are shown in full below.

Overall results


Product malware on demand adware / spyware on demand false positives scan speed proactive detection response times rootkit detection cleaning
AntiVir (Avira) ++ ++ (*1) + ++ + ++ + +
Avast! (Alwil) ++ + + + o + o o
AVG + ++ (*1) + + o o + o
AVK (G Data) ++ ++ o -- + ++ -- -
BitDefender + ++ o - ++ + + o
ClamAV -- -- - -- - ++ -- --
Dr Web o o o o + o + +
eScan + o o - + ++ -- --
eTrust / VET (CA) -- -- ++ o - -- + ++
Fortinet-GW o o -- + ++ + n/a (*2) n/a (*2)
F-Prot (Frisk) + o + + - o o o
F-Secure + o + o ++ + ++ +
Ikarus ++ ++ o + + + o o
K7 Computing -- -- o - - - -- --
Kaspersky + o o - + ++ + +
McAfee + ++ ++ o + o + ++
Microsoft + o ++ o - -- o ++
Nod32 (Eset) + + ++ ++ ++ + + +
Norman o o + - + o o o
Norton (Symantec) + ++ ++ ++ + o ++ ++
Panda + + + + ++ o ++ o
QuickHeal (CAT) - - o o o o - o
Rising o + + o o o o +
Sophos ++ ++ + + ++ + + +
Trend Micro ++ + + + + + ++ +
TrustPort ++ ++ - -- ++ ++ -- --
VBA32 - o o o + o o +
VirusBuster -- -- + o - o o +
WebWasher-GW ++ ++ o ++ ++ ++ n/a (*2) n/a (*2)
ZoneAlarm + o o - + ++ + o

Index
++ = very good > 98% > 98% no FP     < 2 h    
+ = good > 95% > 95% 1 FP     2 - 4 h    
o = satisfactory > 90% > 90% 2 FP     4 - 6 h    
- = poor > 85% > 85% 3 FP     6 - 8 h    
-- = very poor < 85% < 85% > 3 FP     > 8 h    

Notes
(1) the free (personal) edition does not include ad- and spyware detection, so the results would be "--"
(2) not available (this is a gateway product)

Detection rates for malware, adware and spyware


Product Malware samples Adware and Spyware
AntiVir (Avira) 99.3% 99.1%
Avast! (Alwil) 98.8% 97.9%
AVG 96.3% 98.6%
AVK (G Data) 99.9% 99.9%
BitDefender 97.8% 98.8%
ClamAV 84.8% 82.4%
Dr Web 90.4% 92.8%
eScan 96.7% 92.1%
eTrust / VET (CA) 72.1% 56.5%
Fortinet-GW 92.4% 91.2%
F-Prot (Frisk) 96.7% 92.0%
F-Secure 96.8% 93.5%
Ikarus 98.0% 98.8%
K7 Computing 65.5% 59.5%
Kaspersky 97.2% 92.0%
McAfee 95.6% 98.6%
Microsoft 97.8% 91.5%
Nod32 (Eset) 97.8% 96.3%
Norman 92.8% 91.9%
Norton (Symantec) 95.7% 98.6%
Panda 95.6% 95.6%
QuickHeal (CAT) 85.7% 86.7%
Rising 94.1% 95.9%
Sophos 98.1% 98.8%
Trend Micro 98.7% 95.1%
TrustPort 99.6% 99.8%
VBA32 89.9% 92.1%
VirusBuster 76.2% 77.8%
WebWasher-GW 99.9% 99.9%
ZoneAlarm 96.4% 94.5%
Number of samples 1130556 83054

Full testing methodology is here.

Tags: av-test, testing. Posted on 13 March 2008 by Virus Bulletin. 16 comments.

 del.icio.us  digg this! digg this

Quick Links

Poll
Do current laws offer enough protection for ethical ('white-hat') hackers?
Yes, the current laws are fine
No, they prevent responsible disclosure of vulnerabilities
The current laws are too lax, we need to be stricter on hacking
I don't know
Leave a comment
View 4 comments

Jobs Recruit Sidebar

VB2013
VB2013 VB2013 will take place 2 - 4 October 2013 at the Maritim hotel, Berlin, Germany.

Virus Bulletin currently has 227,267 registered users.