Netgear WAX206 vs Xiaomi AX3600

Please advise. I want to replace my RT-N66U with something that has faster Wifi. I've been researching for 2 days and my top choices so far are the X3600 (~80 euro) or the Netgear WAX206 for 55 euro. Price is a secondary factor.

Which one is better for OpenWRT in terms of reliability, transmission speed on 5GHz/Wifi6 (I have 1Gbit LAN and need to get close to that speed with WiFi), and good coverage in a crowded apartment building on the 2.4 band? The 2.4 band will only be used for around 7 IoT devices managed by HomeAssistant.

It will be used either as a dumb AP (currently I have OpenWRT on LXD in HP T620+) or to completely replace the current OpenWRT and use it as an edge device + AP.

My main concern with the WAX206 is the internal antennas, which may impair the wifi range/speed through 1-2 walls.

Dunno about the performance, I use WRX36s, but the WAX is a lot easier to flash.

2 Likes

Technical difficulty is a non issue for me.

Based on a little poll I dropped in another thread, the WRX36 is getting the most love at this point:

I am leaning toward picking up a WRX36 and also a WAX206 for some side-by-side comparison. I'm curious where you land in your decision here, though.

Vote with your wallet I say. I'm not buying anything QCA anymore unless I need to.

Could you elaborate please?

WRX seems the most popular/powerful at the moment. What alternative would you consider?

QCA does not care about FOSS. I'd personally rather purchase hardware that might not be as performant (consider how much of the peak performance you will need?) but in the end does not depend as much on binary blobs as ath10k/ath11k does. That pretty much leaves MediaTek (or outdated 802.11n ath9k solutions). Ath9k was an anomaly where the company grudgingly adapted a FOSS driver that was (ironically) better than their in-house proprietary code. Ath10k and ath11k are QCA trying to profit off the goodwill that ath9k gained them (none thanks to themselves, ironically). While MediaTek is far from perfect either, at least they seem more willing to share details on their hardware and there is a major OpenWrt developer working on the code of the wireless driver. Not a bunch of QCA minions that only do their master's bidding.

People have been arguing that MediaTek 802.11ax solutions aren't as powerful as QCA, but that point is geting moot now that MediaTek ARM 802.11ax solutions are coming to market and getting supported by OpenWrt. Besides that, people lean towards 'let's get the most powerful device I can get' then forget their clients are often mobile devices (smartphones, tablets, laptops) that only rarely come equipped with outsized radios. Most clients have two-stream radios and that's it. You won't benefit from a router with four streams in that case. But it's easy to be fooled by numbers.

My setup (a multi-storey house) is a wired ARM64 box for routing and multiple MediaTek 802.11ax access points. I have a Netgear WAX202 and a TP-Link EAP615-Wall. Those offer two streams maximum, but with good signal I am seeing over 600 Mbps throughput with iperf. So I am a happy camper.

2 Likes

I am leaning towards DL-WRX36, even though the cost to me is more than twice of WAX206 - I'd have to import it from the US.
Reasons:

  • better CPU and more RAM = better base for LXC/Docker containers, maybe even running Home Assistant on it
  • USB enable adding additional storage and Zigbee Rx/Tx device, useful for the above
  • seem so have better signal, throughput and range

Interestingly, OpenWrt community in Poland is most excited about the future prospects of Xiaomi AX3600 once stable OpenWrt is out, calling it the new Archer C7.

Depends on availability I guess, Xiaomi/Mi is pretty big in PL, but not so much in other parts of EU.

Although it's actually cheaper in SE than in PL.

The ax3600 is a well engineered device and works well with OpenWrt, but:

  • ipq8072a (dl-wrx36) is almost twice (yes, not quite, but almost) as fast, 2.2 GHz vs 1.38 GHz on ipq8071a
    that shows, the ax3600 'only' achieves around 600 MBit/s while routing
  • 512 MB RAM (ax3600) is tight with ath11k, the 1 GB in the dl-wrx36 are a considerable improvement
  • one USB3 port (dl-wrx36) vs none
  • lots of small goodies on the dl-wrx (2.5GBASE-T, more 1000BASE-T ports)

Looking at the OEM firmware, the ax3600 would be a failure. While the Chinese firmware is getting (few) updates, that is pretty much unusable by anyone who doesn't read Chinese - the global (English) firmware doesn't get updated at all. While I wouldn't be confident about the security status of either, at least the the global one is completely neglected - no, lets not start about the very chatty Mi service integration.
It's only usable with OpenWrt, but the hardware is well designed (and good passive cooling, very low idle power).

As I have no personal experience with the dl-wrx36 myself, I have to skip looking into it in detail - and I'm certainly basing these comparisons on similar pricing/ availability (while it was in stock at German amazon, the price was 70 EUR for the dl-wrx36 vs 80 EUR for the ax3600).

Resist the urge to (ab-)use your border gateway as general purpose server (docker et al), keep the security exposure as small as possible.

3 Likes

Couldn't agree more.

2 Likes

I have both devices and the AX3600 is currently stored away, mostly because I wasn't able to get it running stable as my main AP/VPN gateway (Xiaomi AX3600 / IPQ8071A ethernet "crashes" sporadically - #7 by NerdyProjects), but so far it seems to me nobody else was able to recreate the issue and I also was not able to reproduce it on my desk, so I don't know what was going on, anyway this was a show stopper for me.

While it was running, I did not have any problems recently with performance or RAM or Wifi (disconnects etc.), but it also wasn't heavily loaded - and ethernet crashed about once per day.

Still, the AX3600 does not have a hardware switch and without the non-standard offloading stuff, running 1 Gbit via the switch loads the main CPU quite a lot! I did not try using unofficial offloading stuff.

The WAX206 is a little bit, but not much, worse regarding Wifi performance/range. It has proven stable so far. I just ordered a second WAX206 for my father to replace his outdated WDR3600 as dumb AP :slight_smile:

In comparison, I trust a "legacy" design like the WAX206 more than a non-vendor-supported specialised, theoretically advanced, platform as the QCA/AX3600 stuff. Having the network processors is great - not having well tested, stable drivers and open source firmware for everything is... a pain in the ass.

I really miss a USB port to have it as a "general-purpose-lay-around-and-grab-for-whatever" access point, that the good old archer C5 has been for me so far. For everything else it is a great upgrade!

Edit: Also, VPN performance on WAX206 is quite good. I don't have exact numbers, but from my power consumption measurements (https://openwrt.org/toh/netgear/wax206#power_consumption), about 150 Mbit/s OpenVPN (or 2-3x that via Wireguard) should be achievable.

Edit 2: For sure you will not be able to reach "Close to 1 Gbit/s" wifi performance through 2 walls. With 160 MHz, depending on the walls, expect 400-600 MBit/s. In the same room, you will reach 1 Gbit/s, yes.

1 Like

This is the most significant reason that I keep feeling compelled to pull the trigger on getting a WAX206. I feel really strongly about supporting MediaTek (as a chipset vendor to other hardware brands) as a monetary "thank you" for being sane and open to the FOSS community. I know the developer of whom you're referring (not sure if we're trying not to mention his name?) and he has been very receptive to some additional MT76 & MT79 troubleshooting that @Brain2000 and I have been working on.

I have loved to see the success of the Linksys E8450 and Belkin RT3200 (of which I own four units presently). I think it's important for the OpenWrt community to really consider what it means to have a chipset manufacturer like MediaTek looking favorably on us. And extending our appreciation to them by purchasing products with their chipsets [hopefully] also extends a positive course with the manufacturers like Netgear, Linksys, Belkin, and others to keep that relationship with MediaTek alive and going.

With that said, I think I have affirmed to myself that the WAX206 is going to be the next WAP I try out. :slight_smile:


On second look, the WAX206 seems to be nearly identical to the RT3200, save for the 2.5Gbps WAN port and different (better?) antenna pattern.

WAX206: OpenWrt support for WAX206 - #6 by Siedrosat
RT3200: Belkin RT3200/Linksys E8450 WiFi AX discussion - #23 by andyrichardson

Anything I'm missing?

1 Like

Yep. Which is why I say: vote with your wallet.

I'm not name-dropping him simply because I figure he gets highlighted enough already. He was very helpful in my (futile) attempts to troubleshoot my MT7613 issues. (Futile because it turned out the mt76 devs got next to no support from MediaTek for that radio which is a wave 2 802.11ac radio - MT7615 on a budget, kind of).

Are you sure? AFAIK the RT3200 has 802.11n 2,4 GHz. WAX206 should have 802.11ax on both radios.

1 Like

Not too sure at all :rofl: I'm a little perplexed by what I'm seeing in the TOH.

WAX206: https://openwrt.org/toh/netgear/wax206#hardware
E8450/RT3200: https://openwrt.org/toh/linksys/e8450#hardware

According to the data there, both have MediaTek MT7622BV for 2.4Ghz. But the TOH for the WAX206 specifically states 4x4 ("MediaTek MT7622BV - 802.11bgn (2.4GHz) 4×4 MIMO").

For 5Ghz, the E8450/RT3200 have a MediaTek MT7915E chipset. But the TOH for the WAX206 seems to have conflicting information:
MediaTek MT7915AN, MediaTek MT7975AN

You're right, 2,4 GHz seems to be 802.11n for both according to WikiDevi as well.

What probably happened is: the MT7975 solution has four streams that can be split over two radios AFAIK (DBDC). The WAX202 e.g. has the same radio, but uses it for both 2,4 GHz and 5 GHz bands, which means each of those has two streams available (MT7621 having no built-in wireless AFAIK). MT7622 has an integrated 2,4 GHz 802.11n radio, so Netgear could use all streams of the MT7975 for 5 GHz, but left you with a 802.11n 2,4 GHz radio.

I believe MT798x SoCs should have more streams available - the TUF-AX4200 e.g. gained support recently and its (also DBDC) radio has 3 streams for 5 GHz and 2 that can be allocated to the 2,4 GHz band or to the 5 GHz one additionally (both AX again).

1 Like

Your assessment of the radio situation makes a lot of sense! Thanks for helping shed more light on that for me (and future readers). :slight_smile:

Ooooh! This is the first I've heard of this unit--thanks for mentioning it! I don't want to take @lockheed's thread any further OT, so I'll go do some more research on the TUF-AX4200.

@lockheed, I hope some of this conversation has been beneficial for your decision. Best of luck whichever direction you go with this!

1 Like

There's more Filogic/MT798x stuff in the pipeline. Take a look at the Makefile for what's already supported now (most devices got added very recently from what I can tell).

1 Like

I can get more than 600 MBit/s at 80 Mhz with AX3600 with two walls (not too far, that's true). So 160 Mhz can give more for sure.

It's install by serial only though.