Replace RT-AC3200 with RT-AX88U or RT-AX86U?

[spacemanspiff]

Well, I came here thinking I should be able to find a pretty straightforward answer and am now completely overwhelmed with all the posts, thoughts and feedback here on the RT-AX86U vs RT-AX88U.

I have an RT-AC3200 that I'm completely happy with except for the fact that there is no future Merlin support. Rolling back to stock firmware is not an option for me as I rely on certain Merlin firmware features.

My dilemma is that Asus Routers are hideously expensive in my current location (New Zealand) and so I'm left with the choice of getting an RT-AX88U for about $100 less than I can find an RT-AX86U.

From what I can tell, the opinion is that the RT-AX86U is the better performing router and the RT-AX88U had some instability issues for users on this forum.

Have the instability issues reported on the RT-AX88U been resolved or are they still present? Is there a compelling reason for me to splash out an extra $100 to get the AX86U?

I don't have any hardware devices that support 2.5gbe so that faster port on the AX86U doesn't seem to offer any value to me. The extra ports on the AX88U would certainly be useful, but I'm far more concerned with performance...especially on running a VPN connection and the performance of my wired connections (I have gigabit fibre and primarily connect via PC directly to my router).

 

[L&LD]

Usually, the RT-AX86U is the lower priced option between these two.

For $100 dollars less, the RT-AX88U would be a good buy.

Won't make it better than the performance that the RT-AX86U will give, but that is what you're trading for keeping $100 more in your wallet.

 

[spacemanspiff]

Usually, the RT-AX86U is the lower priced option between these two.

For $100 dollars less, the RT-AX88U would be a good buy.

Won't make it better than the performance that the RT-AX86U will give, but that is what you're trading for keeping $100 more in your wallet.

What exactly is the better performance the AX-86U is offering vs. the AX-88U. I'm having a little bit of a hard time decoding that from all the various threads/replies/posts on these forums. Just trying to figure out if that performance difference will offer much benefit to me...

 

[thiggins]

Both have been recently tested. You can get an idea of the relative performance vs range by flipping through the "Profile" charts, starting with 2.4 GHz down. The AX86U maintains higher throughput as signal level drops.

Wi-Fi Router Charts

title

www.smallnetbuilder.com

 

[L&LD]

And the latency is noticeably less on the RT-AX86U too.

 

[thiggins]

And the latency is noticeably less on the RT-AX86U too.

Actually, it's the opposite. A higher "score" = lower latency. I had to do this because the Ranker algorithm always sorts from high to low. With latency, lower is better, so I had to invert the number.

The 88U has a 5 GHz 90th percentile score of 65; the 86U score is 45.
Latency = 1/(score/1000). Equivalent latencies are 88U = 15.4 ms, 86U = 22 ms.

Sorry that this is confusing.

 

[spacemanspiff]

so better latency on 88U? I read another review somewhere else earlier that seemed to suggest the 86U was better on the latency front??

 

[leerees]

If you are looking for a bargain, buy the one that's the cheapest. Unless you need the extra features of the AX88u, the difference in performance between them is negligible. Most likely the AX88u will catch up with the AX86u with firmware updates.

I do not believe the test when it is showing latencies in the teens. Even a cheap WiFi router would only add a few ms latency to your connection. The difference in latency between the AX86 and AX88 is not perceivable by a human.

What is evident is that the AX86u has formed a very loyal fanbase on this forum. That's fine, because the AX86u is still one of the best routers you can buy.

 

[L&LD]

Thanks, Tim for those details. I don't doubt your testing or results at all, but in use, the RT-AX86U is the one that feels quicker to me vs. the RT-AX88U.

Even with an RT-AX88U and an RT-AX86U in AiMesh mode (in any combination, main and AiMesh node), inserting the RT-AX88U into the network always resulted in a slower feeling network overall.

That is why when I had the chance to test two RT-AX86U's in AiMesh (v2.0) mode over a 2.5GbE backhaul connection, I was pleasantly surprised that the latency did not increase like it would with the RT-AX88U in the mix.

Needless to say, I sold the RT-AX88U and kept that second RT-AX86U for my network. That is almost 6 months ago, but I don't think anything substantial would have changed on that front today.

 

[thiggins]

so better latency on 88U? I read another review somewhere else earlier that seemed to suggest the 86U was better on the latency front??

Please cite your source, i.e. link to the review.

 

[thiggins]

I do not believe the test when it is showing latencies in the teens.

What do your measurements say?

 

[spacemanspiff]

Please cite your source, i.e. link to the review.

This is the one review that really made me feel like I should spend the extra money on the AX86u instead of saving $100 and getting the AX88U (and losing the ports that would actually be useful to me)

Asus RT-AX86U Review: One of the Best Wi-Fi 6 Routers for the Money

Editor's Choice: If I were buying a router today, I'd get the Asus RT-AX86U. Here's why.

www.cnet.com

In particular, at the end of that review there is a section on latency saying it's the best router the reviewer ever tested and they show this spider graph comparison:

https://www.cnet.com/a/img/kiraAcB3K3I4_Z2s7uK6yq_fMgs=/644x0/filters:gifv()/2020/09/11/3653fd34-d3a9-4f32-b06f-06ad52c31738/latency-graph-gif-asus-rt-ax86u.gif

 

[spacemanspiff]

this was another review that seemed to tip the scale to the 86u https://www.mbreviews.com/asus-rt-ax88u-vs-rt-ax86u/ which seems to suggest it outperforms in most respects

 

[thiggins]

Even with an RT-AX88U and an RT-AX86U in AiMesh mode (in any combination, main and AiMesh node), inserting the RT-AX88U into the network always resulted in a slower feeling network overall.

In both cases, were you running Ethernet backhaul?

If so, do you know where your STA was when it felt slow?

 

[thiggins]

@spacemanspiff Thanks for linking the CNET review. He doesn't say whether his ping test is done from a wired or wireless device. Even though he is pinging the "same" server for his tests, there are many variables in a latency test based on an internet server. Those spikes could be caused temporary internet problems.

At any rate, if you look at the circle at the center of the chart for each device, they're pretty much the same, i.e. the same core latency. What he should say is that the AX86U had the smallest latency variation.

As @leerees notes, you're not going to see the difference between 15 and 22 ms latency. Don't let ie be the deciding factor, or 8 vs. 4 switch ports, either. Gigabit switches are cheap.

 

[thiggins]

Here's a 5 GHz CDF of the AX86U and AX88U

The 86U has wider spread (variation) and higher latency.

Keep in mind my benchmark purposely loads the router heavily and is running both radios at the same time.

 

[leerees]

What do your measurements say?

Like most people, I am here because you've done the hard work for us and that is appreciated. Your test cannot be reproduced because my local WiFi neighborhood is different to yours.

For example, was there any interference from other AP's? Did you mitigate this with OFDMA and beamforming?

Your results are too similar to say the AX86u is better than the AX88u.

Small variances in ping at full load doesn't matter because if the network is planned well it should rarely be at full load. What's important is how the network performs in its most common state. You stated yourself that the variances are insignificant. The "issue" is most people are taking your test as an assumption that the AX86u offers noticeably better performance than the AX88u.

What I would take from your test results is the only deciding factor between these 2 routers is price. Based on the OPs needs, both of these routers are more than adequate.

 

[thiggins]

@leerees Sorry my question wasn't clearer. When you said "I do not believe the test when it is showing latencies in the teens", what did you mean? What values would you believe, higher or lower?

Yes, well designed networks should not have routers running at full load. But keep in mind the multiband test has ONE STA per radio. Granted, they are at maximum signal level. But that condition should not be seriously stressing a router/AP. I could have designed the test to be a lot harder on the router.

 

[L&LD]

In both cases, were you running Ethernet backhaul?

If so, do you know where your STA was when it felt slow?

Yes, wired backhaul in all cases (wireless AiMesh is detrimental to 1Gbps symmetrical ISP connections).

The locations I test at are precisely repeatable.

The network with any client devices (wired or wireless), in regular use / not full load, is noticeably faster with one or two RT-AX86U's than any other Asus routers I have ever used.

 

[Adooni]

yes it is hard if you have very good router like AC3200 and the only reason you need to replaced it because is not support by RMerlin any more. In any other case router you have would be sufficient for next 10 years . This is the reason that I am moving between different software like DD-WRT-tomato- lately RMerlin but in few months I plan to move to UBIQUITI ER-12 router.

I am not sure what are your experiences/ needs with VPN but I tested RT-AX88U at my friend loc and VPN max speed was below 300Mbits, when I tested ER-12 it went 1Gbit, cakeQOS at this Asus router was about 250Mbit and at ER-12 was a bit above 500Mbit.

If your AC3200 was OK for you I would not spend a lot money at new router what think more like replacing it every 5y and go with https://wikidevi.wi-cat.ru/ASUS_RT-AX56U

if you want to have 1Gbit on VPN sry but RT-AX88U will not be even at 1/3 of it.

if you are typical user AX56U will be OK for 1GBit - if you advanced user (QoS, VPN, etc) for 300Mbit above you will need to look for something different than ASUS/Netgear etc.