What's new

Where is the bottleneck? Confused by LAN vs WAN speed discrepancy...

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

DeltaHF

New Around Here
I have an RT-AX86U on stock firmware with a symmetrical gigabit fiber connection and a 14'' MacBook Pro M1 Max.

The MacBook sits about 12' away from the router, unobstructed. macOS reports it is connected via 802.11ax at 1,200 Mbps:

Screen Shot 2022-05-08 at 12.21.41 AM.png


Running various online speed tests, I see ~600Mbps symmetrical speeds via WiFi. Connecting the MacBook to an Ethernet dongle, I can get ~920Mbps.

I fully understand the limitations and overhead associated with wireless connections and I am not surprised by the lower speeds via WiFi.

However... I just copied 100GB+ of files from a PC on my LAN (connected via Ethernet) to my MacBook (still connected via WiFi) and it easily sustained transfer speeds of 950+ Mbps!

Screen Shot 2022-05-08 at 12.15.28 AM.png


This result does not make sense to me. The MacBook and RT-AX86U can clearly talk to each other at 900+ Mbps wirelessly, and the internet connection and speed tests can provide those speeds as well, so where is the bottleneck?

If WiFi is not the bottleneck, what is? Can anyone help me understand what's going on?
 
Well, the semantics of WIFI vs Ethernet are two things but, off the shelf routers are another issue when it comes to throughput.

Going LAN<>LAN though is the test as you saw performance was fine.

WAN - there's a few things that Asus adds to the mix to slow tings down like AIP / NAT acceleration of which anything you don't need besides NAT should be turned off and test again.

If you want to reap the full speed of your line then you need to get beyond using a consumer router and step up to something that goes beyond 1gbps. Also, consider switching the bandwidth for WIFI to 160mhz instead of 80mhz to boost things a bit.
 
Cheers, that's interesting.

I'm not using AiProtect or QoS, but I did have "Traffic Analyzer - Statistic" enabled. Turning that off got me another ~100Mbps in WAN speeds, so we're moving in the right direction. :)

Unfortunately this MacBook Pro's maximum channel bandwidth is 80Mhz and it falls back to 2.4GHz when I specify 160Mhz in the router, so I can't test that.

If I started using another device as the router and converted the AX-86U to a dedicated AP, would I still be running into this issue?
 
I can't say specifically whether or not the performance would increase in AP mode or not. Chances are though it would since all of the underlying gimmicks would be disabled and it basically would be a WIFI switch instead of analyzing packets that traverse to the WAN.

Well, the MBP could be a simple upgrade with an AX210 or AX200 M2 adapter for 160mhz options. The 200 provides 6 and 210 will provide 6E options. Now, looking through Google this seems to be easier said than done. Seems like Apple decided to use the mPCIE format for some models depending on the age of the MBP determine where to go with the WIFI upgrade.

https://www.amazon.com/dp/B07BQ976CR/?tag=snbforums-20

Just an example that gives me something to go off of with more specific info I can narrow down which card / format to recommend. The AX/E line by Intel comes in M2 / mPCIE formats. The best option is to pop the case and snap a pic of it but, a little sleuthing usually works too.
 
Before getting too excited about 160MHz ... do you live anywhere near an airport or weather radar station? If so, you're not going to be able to use 160MHz channels with any reliability. There's just not 160MHz worth of clear bandwidth in the current WiFi 5GHz spectrum; you have to overlap radar frequencies to get that, and the radar service has priority.

WiFi 6e will solve this problem by making additional spectrum available, but that doesn't help you until you have 6e-capable clients and AP.

As for the discrepancy between your WAN and local-on-LAN speed measurements, did you repeat each one multiple times to be sure the results were stable? It could just be that you had more co-channel interference during one test than the other. This'd be more likely if you are not far from your nearest neighbors with 5GHz WiFi gear. (I live in a medium-density urban neighborhood, and I can see my effective WiFi transfer rate varying all over the map depending on what my neighbors are doing. Things could be a lot worse in an apartment block.)
 

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top