Asus RT-AX86U vs RT-AX68U (with Merlin cake SQM)

[tripleg]

Hello,

I am looking to deal with bufferbloat. With my current crappy ISP-issued wifi router/AP, I got a grade of D on the waveform test.

(https://www.waveform.com/tools/bufferbloat)

I intend to buy a router and run SQM on it. I narrowed it down to Asus RT-AX68U and Asus RT-AX86U.

I would prefer the cheaper RT-AX68U and I intend to run Merlin on it and use Cake for SQM. I understand that SQM is CPU heavy and I want to make sure Asus RT-AX68U can handle it.

I am in 600sq/ft apartment. The unit can be placed fairly centrally.

We have 6-8 devices connected, and 2-3 active at once. Typical example of “heavy” use might be Zoom or Discord plus a torrent or two downloading/uploading on one computer and another computer browsing the web or streaming a movie.

The ISP connections is 100/100 Mbsp, but might upgrade to 300/300 Mbps.

Can anyone advise whether the cheaper Asus RT-AX68U can handle this while running SQM with Merlin. Is the 1.8GHz DUAL-CORE processor enough or will I need the 1.8GHz quad-core processor on the Asus RT-AX86U?

 

[Tech9]

I narrowed it down to Asus RT-AX68U and Asus RT-AX86U.

RT-AX68U is End-of-Life model (no more updates), RT-AX86U perhaps has 2 more years of support left.

The cheapest model with newer similar capabilities hardware and expected longer support is RT-AX86U Pro.

I got a grade of D on the waveform test.

What's the real issue? Online bufferbloat tests show worst case scenario at line saturation speeds.

Limit your torrents up/down speed. Your 2-3 active devices will be fine with no extra expense for a new router.

 

[tripleg]

RT-AX68U is End-of-Life model (no more updates), RT-AX86U perhaps has 2 more years of support left.

The cheapest model with newer similar capabilities hardware and expected longer support is RT-AX86U Pro.



What's the real issue? Online bufferbloat tests show worst case scenario at line saturation speeds.

Limit your torrents up/down speed. Your 2-3 active devices will be fine with no extra expense for a new router.

The real issue is zoom and especially discord lag more than I'd like. Like it's noticeable. Especially with screen share.

I tried to give an example with heavies use, but most of the time there is not torrents at all and the issues still happen.

I am in Armenia and usually to video conferencing with the U.S. on zoom. Even zoom calls with people closer to Armenia have a lag though.

On discord it's people from all over the world.

It's hard to judge whether it's because of the ISP, my local network, or the distance. I am not sure how to diagnose where the problem is and if it's solvable. But I want to do what I can from my end to optimize.

Currently I'm using the no-name router issued by the ISP for "free". It's very basic, so I though it might be the issue.

Maybe upgrading to 300 Mbps with a different ISP is more likely to help things?

Thank you for your help.

 

[Tech9]

Your 100/100Mbps is more than enough for audio/video calls. What you are experiencing is perhaps distance related latency and you can't do much about it.

 

[tripleg]

Your 100/100Mbps is more than enough for audio/video calls. What you are experiencing is perhaps distance related latency and you can't do much about it.

I really appreciate the advice to not spend money if it's not gonna help. I am in favor of economy!

Is there a way to diagnose how much of it is the distance and how much of it is my local network? I though the bufferbloat test was part of that diagnosis and the D grade indicated that at least some improvement might be possible with a better router/SQM?

I was also wondering if local WiFi interference doesn't play a role episodically. Like sometimes there is particularly bad lag for a couple of seconds. And when I ping my WiFi router I get 3-6 ms pings but occasionally it spikes to 50ms-150ms.

I wonder if a better Wifi router wouldn't at least eliminate this issue.

 

[Tech9]

I wonder if a better Wifi router wouldn't at least eliminate this issue.

Better Wi-Fi router may help with local issues, but won't help with distance latency.

You can check what you get to different destinations here:

World Ping Test - global ping test

Measure the response (ping) to servers located around the world. The lowest measured value is displayed.

www.meter.net

I though the bufferbloat test was part of that diagnosis

No, online tests are not very accurate and show expected bufferbloat at line saturation speeds. Your audio/video calls don't take even 1/5 of your line speed and if your torrents are not using the bandwidth you perhaps have no bufferbloat at all in normal Internet use situations.

 

[tripleg]

Better Wi-Fi router may help with local issues, but won't help with distance latency.

You can check what you get to different destinations here:

World Ping Test - global ping test

Measure the response (ping) to servers located around the world. The lowest measured value is displayed.

www.meter.net

No, online tests are not very accurate and show expected bufferbloat at line saturation speeds. Your audio/video calls don't take even 1/5 of your line speed and if your torrents are not using the bandwidth you perhaps have no bufferbloat at all in normal Internet use situations.

Thank you.

P.S. It's also kind of funny how in threads from 2021 some people are saying rt-ax68u is "unproven" and it's end of life. lol

 

[dave14305]

Is there a way to diagnose how much of it is the distance and how much of it is my local network?

Use a tool like mtr (on Linux) or WinMTR (Windows) to measure per-hop latency and packet-loss between your network and a remote host of interest.

 

[hancox]

Thank you.

P.S. It's also kind of funny how in threads from 2021 some people are saying rt-ax68u is "unproven" and it's end of life. lol

It was always a piece of junk, despite initial hope (I owned one), and some fervent supporters here. EOL is good riddance, in this case.

 

[Tech9]

and some fervent supporters here

It was never popular. Connectivity issues and firmware delays. See Asuswrt release threads here on SNB Forums, very few people replied there.