Review RT-AC66U vs RT-N66U

[Geraner]

Hi,

I have been using the RT-N66U since January 2012 and wrote also a detailed test report about the RT-N66U back in January. See link bellow.
http://forums.smallnetbuilder.com/showthread.php?t=6446
The RT-N66U I have been using with several open freeware products and ASUS firmware. During the last month using mostly Tomato Shibby.

Now I have been contacted by ASUS Nordic who asked me to test there latest product the ASUS RT-AC66U. I got two devices shipped from them and was testing their new product at the same locations as I was testing the RT-N66U on my tests in January this year.
To test the transfer speed oft he AC standard, one of the RT-AC66U can be used as Wireless Bridge. Because there is no 802.11ac network card existing yet.

To get a better comparison between the RT-N66U and the RT-AC66U I divided the test in several parts.

1. Testing the ASUS RT-N66U with 802.11N with firmware Tomato Shibby 1.28.0000 MIPSR2-100 K26 USB AIO-64K
2. Testing the ASUS RT-N66U with 802.11N with latest ASUS firmware 3.0.0.4.220
3. Testing the ASUS RT-AC66U with 802.11N with latest ASUS firmware 3.0.0.4.220
4. Testing the ASUS RT-AC66U with 802.11AC
5. Testing ASUS RT-N66U with 802.11AC with QOS disabled

Speedtest has been done with the program "LAN Speed Test" version 2.0.8.
http://www.totusoft.com/lanspeed2.html

Test done on 3 locations.
1) My office where the router is located. The laptop is about one meter away from the router.
2) Living room. About 20 meters distance to the router with only 2 thin walls between it.
3) Kitchen. Also about 25 meters distance to the router with 2 thin walls and one thicker wall between.

Measurements have been done with sending/reading of 4 different file sizes.

• 20 MB
• 50 MB
• 100 MB
• 500 MB

Average measured result of all 4 file sizes has been calculated. Results shown bellow are the average results. Measured results shown in Mbps.

The desktop PC which was used to write and read files from via Speed Test was connected to the router via CAT6 cable.
The laptop was used to send and read files from the desktop computer.
WIFI card in the laptop is a Intel Centrino Advanced-N 6205 (2 antenna only!).

The router was set to 40 Mhz when testing N-standard on 2,4 and 5 Ghz band.
RT-AC66 was set to 80 Mhz on the 5 Ghz band for testing AC-Standard.

I also checked the Wirless drivers versions used in all firmware version via telnet command “wl ver”:
RT-N66U Tomato Shibby - Broadcom Wireless Driver 5.100.138.20
RT-N66U ASUS 3.0.0.4.220 - Broadcom Wireless Driver 5.100.138.20
RT-AC66U ASUS 3.0.0.4.220 - Broadcom Wireless Driver 6.30.39.31 (r341183)

Other settings:

• QOS enabled with highest prio to the desktop computer if possible. Otherwise default QOS settings.
• Antivirus has been switched off on both test computers to not affect the test results.

Test results:

Office 2.4 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 99 Mbps
Reading: 71 Mbps

RT-N66U – ASUS firmware
Sending: 99 Mbps
Reading: 124 Mbps

RT-AC66U – ASUS firmware
Sending: 114 Mbps
Reading: 119 Mbps

Living room 2.4 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 71 Mbps
Reading: 79 Mbps

RT-N66U – ASUS firmware
Sending: 85 Mbps
Reading: 112 Mbps

RT-AC66U – ASUS firmware
Sending: 78 Mbps
Reading: 123 Mbps

Kitchen 2.4 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 59 Mbps
Reading: 79 Mbps

RT-N66U – ASUS firmware
Sending: 80 Mbps
Reading: 118 Mbps

RT-AC66U – ASUS firmware
Sending: 78 Mbps
Reading: 94 Mbps

Office 5 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 129 Mbps
Reading: 143 Mbps

RT-N66U – ASUS firmware
Sending: 131 Mbps
Reading: 134 Mbps

RT-AC66U – ASUS firmware
Sending: 122 Mbps
Reading: 146 Mbps

Living room 5 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 125 Mbps
Reading: 142 Mbps

RT-N66U – ASUS firmware
Sending: 124 Mbps
Reading: 128 Mbps

RT-AC66U – ASUS firmware
Sending: 124 Mbps
Reading: 139 Mbps

Kitchen 5 Ghz 802.11-N

RT-N66U - Tomato Shibby
Sending: 106 Mbps
Reading: 107 Mbps

RT-N66U – ASUS firmware
Sending: 101 Mbps
Reading: 117 Mbps

RT-AC66U – ASUS firmware
Sending: 101 Mbps
Reading: 133 Mbps

Office 5 Ghz 802.11-AC

RT-AC66U – ASUS firmware
QoS enabled

Sending: 217 Mbps
Reading: 264 Mbps

RT-AC66U – ASUS firmware
QoS disabled

Sending: 312 Mbps
Reading: 322 Mbps

Livingroom 5 Ghz 802.11-AC

RT-AC66U – ASUS firmware
QoS enabled

Sending: 217 Mbps
Reading: 264 Mbps

RT-AC66U – ASUS firmware
QoS disabled

Sending: 317 Mbps
Reading: 289 Mbps

Kitchen 5 Ghz 802.11-AC

RT-AC66U – ASUS firmware
QoS enabled

Sending: 216 Mbps
Reading: 268 Mbps

RT-AC66U – ASUS firmware
QoS disabled

Sending: 272 Mbps
Reading: 289 Mbps

Cable connected through router

RT-N66U - Tomato Shibby
Sending: 703 Mbps
Reading: 492 Mbps

RT-N66U – ASUS firmware
Sending: 704 Mbps
Reading: 511 Mbps

RT-AC66U – ASUS firmware
Sending: 714 Mbps
Reading: 538 Mbps


Conclusions so far:

The 2,4 GHz N speed of the RT-N66U with Tomato Shibby firmware is slower than with ASUS firmware installed on the product. The 5 GHz N speed on both firmware versions are about the same. (+/- 10 Mbps on some locations)

Comparing 2,4 GHz N and 5 GHz N speed on the RT-N66U and RT-AC66U shows no real differences.
Only the 2,4 GHz N speed of the RT-AC66U is about 10 Mbps slower at the location most far away from the router (kitchen). But that’s not much compared to the total speed and the average user should not notice any difference here.
I have seen some people complaining about that the 2,4 GHz of the RT-AC66U is not as good as on the RT-N66U. According to my test results, I can’t confirm that.

Comparing 802.11N with 802.11ac shows a big difference in speed.
With QoS enabled the speed was identical at all locations in my flat. (80 – 140 Mbps vs 217 - 267 Mbps)
When disabling QoS in the RT-AC66U, the transfer speed could be increased by additionally 100 Mbps in sending and 50 Mbps in reading to a speed of 270 to 322 Mbps depending on the test location.

ASUS advertising is telling us that 802.11AC is about 3x faster than 802.11N, which I can agree as long as the user has QoS disabled on the router.
The advertising and is also telling us:
“Compared with current 802.11n, the RT-AC66U delivers up to 1.3Gbps at 5GHz, which is 3X faster than 802.11n!” written in small letters under it: “*Based on theoretical maximum speeds in 5GHz transmission,...”
http://www.asus.com/Networks/Wireless_Routers/RTAC66U/#overview

Even if the theoretically transfer speed used in the product advertising never can be reached, the 802.11AC performance is in my point of view is impressive.

Now we users are just waiting for the manufactures to release 802.11ac wireless cards as USB and build into new laptops, to start taking advantage of the 3x faster wifi speed that comes with this new standard.

Many thanks to ASUS Nordic to give me the opportunity to test their new RT-AC66U.

Regards

 

[RMerlin]

The performance difference with QoS disabled is due to the fact that CTF is only enabled when QoS is disabled. You can probably reproduce the same performance drop by going to LAN, and setting "Disable HW Acceleration" to "Yes", which will force the router into not using CTF.

That also explains part of the performance difference with Tomato, which does not have that CTF module.

 

[Geraner]

Thanks for the comment RMerlin.
What stands CTF for? Can you add more information please? Thanks.

/Geraner

 

[RMerlin]

Thanks for the comment RMerlin.
What stands CTF for? Can you add more information please? Thanks.

/Geraner

Cut-Through Forwarding. It is essentially a proprietary (closed-source) kernel module from Broadcom that speeds up packet processing by the kernel. It has a few limitations however, which means it has to be disabled by the firmware whenever you enable QoS. In the past it also had to be disabled when using port forwarding, which made CTF more or less a "benchmark review booster" essentially. This is no longer the case.

 

[dopefish]

Thank you very much for this excellent review.

I can't decide between RT-AC66U and RT-N66U.
The RT-AC66U costs 50 Euros more (170 to 120 - Amazon Germany).

What do you think?

Is it worth investing 50 Euros more into 802.11ac?

 

[RMerlin]

Thank you very much for this excellent review.

I can't decide between RT-AC66U and RT-N66U.
The RT-AC66U costs 50 Euros more (170 to 120 - Amazon Germany).

What do you think?

Is it worth investing 50 Euros more into 802.11ac?

Not for that big of a difference, no. You most likely won't get any 802.11ac network interfaces until next year. so might as well wait until non-draft and less expensive products come out.

 

[Geraner]

Not for that big of a difference, no. You most likely won't get any 802.11ac network interfaces until next year. so might as well wait until non-draft and less expensive products come out.

Can only agree with what RMerlin worte.

 

[dopefish]

Not for that big of a difference, no. You most likely won't get any 802.11ac network interfaces until next year. so might as well wait until non-draft and less expensive products come out.

Thanks! I just ordered the rt-n66u (118Euros).

Btw: One of the reasons i bought this device is your firmware RMerlin. Asus should give you some credit for that! ;-)

 

[Bearxor]

Same question. Picking up a new router tonight at best buy. N66 is $180, ac66 is $200. Worth $20?

I understand I probably won't have any ac clients for several years, but the newness of the ac66 and the 128MB of flash are really calling me. Worth $20 you guys think?

 

[Pierino]

Same question. Picking up a new router tonight at best buy. N66 is $180, ac66 is $200. Worth $20?

I understand I probably won't have any ac clients for several years, but the newness of the ac66 and the 128MB of flash are really calling me. Worth $20 you guys think?

I'm wondering the same thing. I can return my n66u and pick up the ac66u for about $20 more.
I'm just concerned about the fact that all the reviews say the 2.4g range is not as good as the n66u. Including the review from our own Mr. Higgins, although I'm wondering if that has changed with subsequent firmware releases.

 

[RMerlin]

I'm wondering if that has changed with subsequent firmware releases.

There's still some improvement that can be made to the driver, since this is a fairly new chipset.

 

[Pierino]

You have both Merlin, do you notice a difference in range on the 2.4ghz band?

 

[RMerlin]

You have both Merlin, do you notice a difference in range on the 2.4ghz band?

I only did a quick test a few weeks ago, and there was no measurable differences for me. I had like a 3-4 dB variation maybe, testing both routers in the next room.

 

[Bearxor]

I ended up picking up the N66R, but it was just a little too flaky with my device connections. I took bit back.. Back to the drawing board.

 

[Ggamer]

I too have same problem with RT-N66U, just would not be stable for me, I have many devices and two 8 port switches and a server and access points. unforchantly I had to return the RT-N66u

Good news i perchased the RT-AC66u and with its newest firmware its great no problems at all. It's routing is incredibly stable and fast the best I ever used and i used a lot. It's a keeper. Now I use AP's running off it and not using the wireless on it so can't comment on that.

So its handling everything flawless and i finally couldnt be happier after 20 years of going crazy. Lol

Oh now not sure if it makes a difference or how it would effect the network or router though I should mention I also updated the linux servers Ethernet from Realtek to intel at the same time as the RT-AC66U install, also turned on jumbo frames on router.
I mention this cause I had a few routers that when I connected my server to them the router would just freeze up, namely ASUS and a few others, :/ I kid you not, also many routers did not like running AP'S from them they would get flaky

 

[randomName]

Cut-Through Forwarding. It is essentially a proprietary (closed-source) kernel module from Broadcom that speeds up packet processing by the kernel. It has a few limitations however, which means it has to be disabled by the firmware whenever you enable QoS. In the past it also had to be disabled when using port forwarding, which made CTF more or less a "benchmark review booster" essentially. This is no longer the case.

Does your firmware offer explanations like this? I didn't know about CTF

 

[RMerlin]

Does your firmware offer explanations like this? I didn't know about CTF

No. This is the kind of knowledge that you generally have to dig through Google.

 

[Poseidon]

Returned my N66U today and picked up the AC66U instead. I could NEVER get the 5GHz to work w/ the N66U (lose connection when away from router by 15-20 feet) - but 5GHz is working great thus far on the AC66U. Not a single drop-out even in different rooms of the house.

Also I haven't lost connection between the router and modem when IP address renews (major issue w/ N66U). This is the reason WHY i ended up returning the N66U. Drove me crazy w/ daily connection loss.

Night and day difference for ME.

 

[MonkY]

Thanks for the comparison! As you have tested both RT-AC66U & RT-N66U, can you please tell us how is the heating issue? Did Asus solved the problem (although the internal design seems the same). Is the new AC66U getting the same heat amount as N66U does? My N66U really worries me, as it's getting really, really hot!

 

[RMerlin]

Thanks for the comparison! As you have tested both RT-AC66U & RT-N66U, can you please tell us how is the heating issue? Did Asus solved the problem (although the internal design seems the same). Is the new AC66U getting the same heat amount as N66U does? My N66U really worries me, as it's getting really, really hot!

There is nothing wrong with the amount of heat generated, the hardware was designed for this.

My former WRT320N was actually running hotter than this, leaving the wooden shelf on which it sat quite warm to the touch. My cablemodem also runs much warmer than my RT-AC66U.