What's new

ASUS ZenWiFi AX6600 upload speeds consistently slower than download speeds

  • 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!

starbelly

Occasional Visitor
I have a CenturyLink gigabit fiber service that I use with my ASUS ZenWiFi AX6600 and consistently get upload speeds that are markedly slower compared to my download speeds over wifi. On average, I'll get 500-600mbps for downloads and 200-300 mbps for uploads over wifi. However, when I connect via ethernet to the network switch that is connected to a LAN port in my AX6600, I get symmetrical gigabit upload and download speeds.

What could be contributing to the difference in download vs upload speeds over wifi? Is there something in the configuration of the AX6600 that may be causing this?

Any help would be appreciated!
 
did you search the threads here ?
I recall several on this issue and what works for some or not.
 
Not only on XT8. I am seeing the same on all my ET12. But my XD6/XD6S are full speed up/down on wireless.
 
It's entirely likely that this is the fault of the wireless client gear, which you didn't even say what it is ...
 
It's entirely likely that this is the fault of the wireless client gear, which you didn't even say what it is ...
Well, on my 2 pair of ET12 (so 4 units). I did try on my ZFold4/ZFold3/ROG Flow X13 2022 (Mediatek)/ROG Flow Z13 (Intel AX211) on WIFI6E/WIFI6.
And Surface Laptop Studio/Galaxy Tab S8 Plus/ZFlip 4/ZFlip 3/Asus Zenbook Pro 16X OLED on WIFI6 (Intel AX211) on 5GHz.

Wireless upload throughput drop is pretty consistent across all these devices. And if I just use my fleet of XD6/XD6S instead on the ET12, no such wireless upload throughput drop is experienced.

Don't know if that's enough devices to demonstrate the problem.......
 
My temporary fix for the problem is to only use 1G LAN ports on the ET12 for uplink to my main multigig-switch, or to use a 1G switch in between ET12 and my multigig switch to force the ET12 2.5G LAN ports into 1G link speed. And my upload throughput is back to close to 95% of my 1G fiber WAN.

I will have to live with either the ET12 unit leds turning RED (because it can't reach internet on the WAN 2.5G ports), or turn the leds off...... And obviously no 2.5Gbps speed to LAN. Additional side effects are that AiMesh is kind of pointless as the other ET12 nodes come up as disconnected and cannot be managed via the primary ET12 without uplink connected via one of the 2.5G ports........
 
Last edited:
It's entirely likely that this is the fault of the wireless client gear, which you didn't even say what it is ...
Hey there! Could you clarify what you mean by this? All In have are 3 ASUS ZenWiFi AX6600 units. The discrepancy with upload speeds only exists on wifi; I get the full gig up and down when I connect via ethernet to the network switch connected to the LAN port of the AX6600 that is connected to the ONT.
 
While that upload speed does sound a little low, it's plenty for most purposes...unless you have an internet server that's connected to the internet via wi-fi maybe *smile*. We also have symmetric fiber here, and I've never seen the upload speed be as high as the download speed with wi-fi, while, as you said, it is symmetric for wired clients. The further away the source of the wi-fi is, the less symmetric it seems to get. Given that before I had fiber I was on a 150/5 broadband speed tier with inadequate upload capacity, I've never found myself short of upload speed since getting fiber *smile*.

Currently we are on a 500/500Mbps speed tier, provisioned at about 620/620Mbps. While I see about 580 - 600Mbps download on wi-fi, upload can be about 300 - 450Mbps...certainly not symmetric, but way more than adequate. I've seen this with fiber on a wide variety of routers, mesh systems, and client devices here, so I just ignore it. Why look for problems where things are working well?
 
Hey there! Could you clarify what you mean by this? All In have are 3 ASUS ZenWiFi AX6600 units. The discrepancy with upload speeds only exists on wifi; I get the full gig up and down when I connect via ethernet to the network switch connected to the LAN port of the AX6600 that is connected to the ONT.
Right, therefore there is no problem with the fiber service, nor with the base AX6600's connection to it. The problem lies in the wireless behavior, which is as much dependent on the client devices' behavior as on the AX6600's. Some possibilities:

* Client's raw Tx rate is less than the router's. This'd likely stem from the client having lower transmit power than the router. If the speed discrepancy changes with the client's distance from the router, I'd suspect this first.

* Client is using fewer spatial streams to send than the router is. Most client gear that I've heard of has the same number-of-streams capability for send and receive --- most current gear is "2x2" meaning it can do 2 streams both ways. But just because it can do it doesn't mean it is doing it. I'm not enough of an expert to know what controls that.

* Client is not optimized to send continuously. Portable (laptop, phone, ...) clients are likely set up with a strong eye to battery life, which discourages the manufacturers from configuring them to blast full wifi transmit power continuously. I'm fairly sure that a lot of the ping-response variability I see with my Apple laptops is down to the laptops shutting down their wifi gear off-and-on (for just tens of milliseconds at a time, but nonetheless off) to save power.

* Since you mention having a mesh setup: much depends on whether the client is connected to the base station or a remote node. If it's connected to a remote, then the backhaul link is going to make a dent in your performance.

Depending on what client gear you're using, this low-level behavior might or might not be configurable, or even visible to you as a setting. But it'll affect the behavior you see when trying to benchmark maximum upload/download speed. If the manufacturer did a good job of making their tradeoffs, it won't much affect ordinary real-world use, though. (@RogerSC makes more or less that same point nearby.)
 

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