Real world speed/reliability/distance differences? Worth upgrade AC86U to AX6000 and SB8611? Testing....

[weaverinva]

I appreciate all the advice offered here. I have read many many posts by regulars here...thank you for offering your service and assistance to the community.

With that said, I have xfinity 1200/40 service with SB8600 (1Gb) with ac86u, two older netgear ac routers set as AP via wired backhaul, and 1gb infrastructure (switches). I get 941mbps to my 4 wired desktops. I have 5 people (two college students, one high school, and two adults and I work from home as an engineer. I got the gigabit service because I wanted the 35(40)Mb up because the next lower was 25. We have roughly 18 devices in the 2000sq ft house (main and basement each).

I upgraded from n66u in 2020 to an AC86u and the Motorola SB8600 and things are mostly well. However, I cannot take advantage of the over provisioned gigabit service. Xfinity with the SB8611 that I have been testing can do 1366 down wired and 885 down via an AX6000.

I am a prudent spender.....and I spent $765 to try my experiment......SB8611 with 2.5Gbps port to AX6000 with 2 ports. Internal 2.5Gb will go to a new TrendNet 2.5Gbps 16 hub for the 3 wired computers and then for one remote basement corner, I will put in the ac86u, and one for outside usage at a back window. I understand the loss of speed but pragmatic. AX6000 would cover the central house. Most of my wifi devices are ac wifi5. The AX6000 seems to provide 20-100% better download speed for the same distance on wifi5 devices.

Though with this investment, unless transferring large files, it does not seem this upgrade will make much of a real world noticable difference? Does this extra downspeed make much difference?

Any comments on this situation or considerations? Debating to keep 'my investment' or send the whole kit and kabootle back and wait 2-3 years for wifi7 and DOCSIS 4.0.

Thank you in advance for any advice. Peace.

 

[L&LD]

If that download speed is consistently faster, that means the latency is inversely lower too.

I don't know if $765 is worth it to you for a lower latency experience, but it would be for me. Particularly as you upgrade equipment over time to new standards (as we all do, as old devices are replaced or new niche devices are deemed necessary for us in the future).

I don't push my 1Gbps symmetrical ISP speeds more than a few times a week. But I enjoy the low latency network I have all the time (RT-AX86U main with GT-AX6000 AiMesh node in wired (2.5GbE) backhaul mode).

 

[Tech Junky]

If you want the over provisioning you need to bond 2 ports together and have a device on the other side that can do so as well.

I used an mb8600 which has 4 ports and lacp of multiple ports and a diy router to get the extra bandwidth.

The best option though would be a CM that has a 2.5 port and a router with a 2.5 wan or USB Ethernet adapter. Unless you want to build your own and put in a multi port NIC and setup bonding.

 

[Tech9]

Though with this investment, unless transferring large files, it does not seem this upgrade will make much of a real world noticable difference?

Correct. You have invested in better speedtest numbers basically.

 

[weaverinva]

Good points. Thank you both. My current MB8600 is a single gigabit wan port. No LACP possible. Thus the MB8611 for testing.

The 20-100% faster wifi speeds seem and the slight distance benefits of the AX6000 lean me that direction but I suppose if I replaced the primary router with an ax86u vs an gt-ax6000... why would I just replace the router for roughly $250-350 and not do the rest of the 2.5Gbps infrastructure.....but of course it will all be cheaper in the future but in how many years.....and the enjoyment across that time is worth something. I work from home all the time now......on Zoom & teams meetings, teams channels, Sharepoint drives, Onedrive, and sending many things via Outlook as well.

However.....941mbps versus 1366 is not really going to be noticable likey because both sides of servers need to come up to speed and use multiple streams......That is what Ookla speedtest uses.

A better router would deliver faster app updates to the 6 phones in my household but really so what.

Just thinking out loud.....

Peace.

 

[weaverinva]

Correct. You have invested in better speedtest numbers basically.

Right. That is what an older WSJ article says......it measured many streams and excess beyond 300Mbps does not make much difference.

The Truth About Faster Internet: It’s Not Worth It

WSJ testing shows typical U.S. households don’t use most of their bandwidth while streaming and get marginal gains from upgrading speeds

www.wsj.com

Why did you invest in the AX-86u? Of course a great router.....

 

[weaverinva]

You know....perhaps I invest in a used AX-86u and get better wifi numbers with wifi6 and forget the rest. Save me $600 and I get the better central wifi speeds for wifi 6 clients. Then all I am missing would be the difference between 941 and 1366 for 3 wired PCs and I really don't use that many times.

Though with the AX6000 and other infrastructure, I could deliver true 450Mbps to 3 computers simultaneously.

Hard time for me to justify just better speed test numbers.

I am not a torrenter at this point. The GT-AX6000 provides the ability to VPN specific IPs via the router for different services etc. Wow, that's fancy.

 

[Tech9]

Why did you invest in the AX-86u?

I'm not using it for Internet access. Purchased one to see what it is. It's a toy. Will be perhaps donated some time after.

 

[Tech Junky]

because I wanted the 35(40)Mb up

Now, you might consider dropping the speed plan to $50/mo and adding a second connection through T-Mobile or Verizon. Uploads with TM are usually at least 70mbps and top out for me at 100 under ideal conditions at night usually. You might even get a credit through work for remote work.

 

[eightiescalling]

Right. That is what an older WSJ article says......it measured many streams and excess beyond 300Mbps does not make much difference.

The Truth About Faster Internet: It’s Not Worth It

WSJ testing shows typical U.S. households don’t use most of their bandwidth while streaming and get marginal gains from upgrading speeds

www.wsj.com

Why did you invest in the AX-86u? Of course a great router.....

I'd say that article is spot on when it comes to general household usage. Interesting gamers only get a brief reference as many people in the UK I know on gig services do it specifically because of the demands game updates have at the point the gamer wants to play - you don't want to wait half an hour when you've just got in from work.

What it doesn't touch on though is that speed vs capacity in the ISPs network are 2 different things.

I'd pretty much guarantee your ISP doesn't have capacity to serve the high speed users they have en-masse. They do however need the subscription dollars from those users to get enough capacity for the more general levels discussed here and in the article.

 

[Tech9]

Most ISPs offer Gigabit connections just because they know very well most users won't use even 10% of this speed 90% of the time. As a result more users pay higher monthly fees and some even go ahead and spend more money on network upgrades for no added benefits. The business is good.

 

[Tech Junky]

use even 10% of this speed 90% of the time

And in the enterprise world we scale as needed and provision 2X the bandwidth needed on a single link for fail over purposes. We also diversify the connections coming in from different access directions into the building to prevent a single cable cut from taking both connections down.

In some instances you can buy links where they scale based on demand but, they also charge more for those bursts above your subscribed level.

The trunks from the neighborhood coming back into the core facility and aggregate the traffic are provisioned appropriately. It's always a task though for performance and planning to constantly remeasure the capacity vs utilization. Forecasting can only do so much as things are constantly changing in user data profiles. The routers / switches through can easily be updated with card swaps to provide the bandwidth needed. Lining up additional feeds though from providers takes time to provision unless they have additional fiber strands not in use to the location that they can simply light up on demand. It depends on how densely populated the area is as to how many strands they will allocate to a conduit coming into a specific building / area.

It's all a game of numbers though to meet demand but keep the costs low. Subscriber fees get split among different aspects of the business but, general the biggest portion is put into maintaining / increasing the network reliability and capacity to bring more people online and keep expanding. If you don't grow the network and keep people happy you stagnate and start losing cash flow. Something a few companies should do and just fade into history,

 

[Tech9]

Subscriber fees get split among different aspects of the business but, general the biggest portion is put into maintaining / increasing the network reliability and capacity to bring more people online and keep expanding.

You have to see the profits our monopolies make in Canada. We pay one of the highest rates for communications.

 

[Tech Junky]

You have to see the profits our monopolies make in Canada. We pay one of the highest rates for communications.

Welcome to the British colony of taxes. On the flip side you pay less for medication to relieve the headaches and stress. When you think about the monarch estates 1T for Charles and 750B for William it makes sense. They each want their cut.

 

[Tech9]

Welcome to the British colony of taxes.

We are talking about greed and unfair business practices here, crushing every possible competition in early stages and in any possible way. The three communications monsters generate >50 billion revenue in a country with <40 million population. I'm only temporary here.

 

[weaverinva]

I'm not using it for Internet access. Purchased one to see what it is. It's a toy. Will be perhaps donated some time after.

@Tech9 do you find the new ax86u toy valuable? Any better speeds? Or distance? That are helpful to you? Curious of your observations.

 

[weaverinva]

Thank you for all the good replies here. I will either use two ac86u and may add a central ax86u.... We'll see. Don't really need the multi gig yet....... Especially if Wifi6 is only 40% faster that wifi5 according to Dong knows tech.... Good articles.

Gigabit Internet and that Cool 1Gbps Mark | Dong Knows Tech

You just upgraded to Gigabit Internet, or even faster broadband, and suddenly have that burning bandwidth question? This post will explain it all.

dongknows.com

So true about businesses provisioning progressively and for redundancy.

Peace

 

[Tech9]

@Tech9 do you find the new ax86u toy valuable? Any better speeds? Or distance? That are helpful to you? Curious of your observations.

I purchased this one only to see how good it is, based on forum feedback. I never had any intentions to use it on my network. It's a good enough home router for the money, but with more advertising involved than real useable features, especially if you already know how things work on business class hardware. It depends what do you compare it to and what do you want to use it for. If you already have AC86U and mostly AC class clients, you won't be impressed much by wireless speed and range. If you want to run more VPN clients on it or heavier custom scripts - it's better with 2x CPU cores and 2x RAM. Asuswrt-Merlin offers unique ecosystem in consumer products and I wanted to have some good hardware to run it on. If you check my profile you'll see I had quite a bit Asus routers in my collection, but most of them were donated to charity. This is where AX86U is heading after I finish playing with it. In my eyes it's a plastic toy with $50 hardware inside, but for someone else it may be a great network upgrade with familiar user friendly UI.

 

[Tech Junky]

Dong

Decent site for some things but, not comprehensive.

Understanding Wi-Fi 4/5/6/6E/7 (802.11 n/ac/ax/be)

 

[L&LD]

That article hides behind a paywall. But I can be pretty sure it ignores what I've stated in post 2 above.

Being a two + year old article doesn't help its stance today either.

What I know is that with a higher-end package, many other benefits are extended to the customer. No monthly limits, for one. And many others depending on the other services you're using from the same provider.

The arguments provided above are the same for other purchases such as sports/luxury cars, etc. In other words, not with merit.

If the small price jump (I've seen the price reduced for the faster package too, many times) is affordable to any specific person, the benefits are real. Regardless of the positions taken above by many.

Having a service that reduces your waiting time is worth a lot to me. And worth it to many others too who give it a fair test too. In my case, the reduced latency is worth more than a maximum upload speed (above ~750Mbps, for my use cases).

Just like the feel of a specific sports car can be worth 10x the money to a specific person, and they don't need to use the performance offered 100% of the time either... so is faster internet the same too. Except it doesn't cost 10x, you can max it out daily, if you wish, but the feel of the network is much more seamless with a lower latency connection.

When I upgraded my network with 2x RT-AX86Us with 2.5GbE backhaul between them and (with a couple of switches) to all my wired capable devices too (including my laptop), it was a game changer.

Just like moving from network hubs to 10MbE switches was. And then again to 100MbE too.

You can use many words justifying not getting a faster base network, but that doesn't make those reasons real.