Will A New AC Router Bring You Wi-Fi Joy?

[thiggins]

So no comments on the article from the peanut gallery? No one want to comment on how poorly the AC88U did?

 

[Bodders]

Funnily enough I have just registered with the forum as a result of that article . The aim of the article was, I believe, to clarify whether the latest WiFi models were worthwhile upgrading to from AC1900 devices. All things being equal (5GHz is 5GHz no matter how many bands you have) it's hardly surprising that there was no benefit and with early drivers perhaps there is even a 'dis-benefit'.
As I think you have pointed out elsewhere in the forum, early adopters are nearly always disappointed.

One question I do have is could MU-MIMO have any benefit if used in a wireless bridge? I currently use 2* RT-AC66U (i.e. AC1750) to connect my office back to the house over 5GHz. The distance is about 10m. Link speed varies from 702-867Mbps and round trip latency is about 1.7ms

 

[Charly]

So no comments on the article from the peanut gallery? No one want to comment on how poorly the AC88U did?

Thank you Tim. I am a new member, and I love to see these more scientific tests to assess hardware performance.
I see your point. I wished I've read your 2015 guide before I rushed to purchase a new router .
I am using the AC88U (got a combo through New Egg), and I am ecstatic about its performance, although
I might have gotten that from any other AC router though.

Still, will you be able to retest the AC88U with the new firmware (your test seems to be running with an older one)? It has the same components as the ASUS AC5300 (which you tested with the new firmware), so is strange to see the gap between the two.
Thanks!

 

[Deleted member 22229]

http://www.smallnetbuilder.com/wire...32896-will-a-new-ac-router-bring-you-wifi-joy

Thanks interesting read..

Tim you mentioned that the 5ghz lower and upper bands now use the same higher power allowed by the fcc in 2014 and all routers you tested reflect that change. That said are there any advantages to the lower or higher channels now that they transmit the same power levels ?

 

[thiggins]

That said are there any advantages to the lower or higher channels now that they transmit the same power levels ?

None that I know of. The reason you see a difference in the "Tri-band" routers is one of the 5 GHz radios shares external dual-band antennas and the other has its own internal 5 GHz antennas.

 

[thiggins]

Still, will you be able to retest the AC88U with the new firmware (your test seems to be running with an older one)?

I checked all sites for new firmware before testing.
The RT-AC5300 and AC88U will have full reviews.

 

[thiggins]

One question I do have is could MU-MIMO have any benefit if used in a wireless bridge?

No. MU-MIMO enables higher total throughput use for multiple devices for AP/router transmit only. It does this by letting up to three devices share the same transmit slot air time.

You need two devices to achieve any benefit. With current 4x4 designs, optimum throughput gain if obtained with three devices. Throughput gain falls off with 4 devices and up.

 

[sfx2000]

So no comments on the article from the peanut gallery? No one want to comment on how poorly the AC88U did?

Saw the article, and it pretty much confirms my thoughts on the X-Stream architecture as having challenges... the devil is in the details on the RF front end, and having two high power transmitters in the same band (different channels but still same band) has near-field impact that is hard to solve...

The 88U was a bit odd, but perhaps due to early firmware?

 

[sfx2000]

BTW - the AC680 solution for the MotoX - if Moto is following their past actions, it's a Qualcomm solution, WiFi MAC/BB on the Snapdragon, and a Qualcomm WCN RF solution for BT/WiFi/FM - it's not a discrete WiFi chip, but much more integrated into the Qualcomm SoC (lower cost) - model number WCN3680 - It's a single stream Wave 1 Su-MIMO 802.11ac solution supporting up to MCS-9, general feedback is that it's good for about 200 Mbps on the 5GHz side..

 

[RMerlin]

I checked all sites for new firmware before testing.
The RT-AC5300 and AC88U will have full reviews.

You list the RT-AC88U as running firmware 378_85xx, which was a pre-release firmware flashed on the first batches sold. The first time you configure it, it should have automatically updated to the release 380_xxx firmware (the latest available from the Asus website is 3.0.0.4.380_858). I wonder if that might explain why the RT-AC88U didn't match the performance of the RT-AC5300 despite having the same SoC, and most likely very similar software configuration, as 378 still used an older driver (and if it's like my own sample, that firmware still had a lot of debugging enabled in it).

 

[Just Checking]

So no comments on the article from the peanut gallery? No one want to comment on how poorly the AC88U did?

Tim, Thanks for doing this research. I am always interested in the performance of the "Latest and Greatest" routers in "real world" environments.

This article seems to verify and extend my own "in the wild" testing that showed no real performance gains in wireless downlink or uplink between Asus RT-AC66, RT-AC68, RT-AC87 and NetGear R-7000 routers. Now it is showing that the newer, more expensive routers have no better WiFi performance either with the client devices most of us have to deal with.

For under US$10 I bet that I can change out an external antenna on the Asus RT-AC66, RT-AC68 or NetGear R-7000 and improve the 2.4GHz uplink and downlink performance to match, or beat, that of the newer routers when it comes to "real world" testing where the client device is more than a few meters away from the router.

 

[simonG]

Tim, cant wait to read your full review on the RT AC88U. I must say I was about to get an 88U and your article stopped me in my tracks. I roughly expected the wifi performance to be similar to existing AC1900 units, within different manufacturers variations - there is no way I expected that big a decrease. That said Rmerlin's comments about drivers give some confidence the 88U can deliver on the promise of it's specs. Understandably you couldn't test a firmware which wasn't there when you tried, but it would be great if the full test has the latest firmware. Huge thanks for all your good work on this.

 

[Puppa]

@thiggins $495 for the RT-AC5300? I've already got sticker shock at $399. Is this Canadian pricing? Or maybe you're just reporting the "going rate" thanks to all the techno-scalpers that have invaded the game for this router on Amazon?

Regarding the lower/higher bands, it seems that my cell phones favor the lower bands, always reporting better wifi signal, but I'm sure a lot of environmental variables are in play.

 

[System Error Message]

i think part of the problem comes from the fact that theres only 1Gb/s from wire from all 8 ports to CPU, 1Gb/s from 4 ports to the other 4, so it is poorly designed like the AC87U. Although your tests were done with a 1x1 client it is unclear if wifi of these newer than AC1900/3200 routers have wifi connected to CPU or switch and could have serious problem whenever an internal bus becomes a bottleneck.

It would be great if you could repeat the tests using newer firmware and showing how apparent this bottleneck is such as wifi clients maxing out wire. On the ASUS AC3200 the wifi is connected to switch so you can have 2 or 3 wired clients/bonded to fully serve wifi but not im sure on the newer routers and not all manufacturers use this approach. In order to fully use the wifi bandwidth you would need to spam multiple synthetic benchmarks at the same time. You can also do LAN transfers on the AC88U between the 2 switches using more than 1 port on both switches to show how bad the design is because im sure people expect to use it for NAS, file transfers and operations that would want to transfer more than 1Gb/s sometimes between switches.

1Gb/s link to CPU isnt a problem if wifi is slower than 1Gb/s to CPU or it is only going to WAN.

 

[stevech]

WiFi RF front ends do not have a channel-width bandpass filter. Would have to be tunable. Way too expenensive. Especially if it had sharp channel edge skirts - and that causes group delay which wrecks the transmitted signal's EVM/rho (quality).

So, operating a transmitter within the antennas' near field yields, of course, lots of self-jamming - if there are receivers working independently in the same band.
Multi-sector system like celluar and 802.16e use time-synchronized MAC layers so all transmissions, no matter the channel, are at the same time.
Two flaws in this: It only works well with TDMA systems. Not worth it for 802.11's CDMA.

 

[Nullity]

None that I know of. The reason you see a difference in the "Tri-band" routers is one of the 5 GHz radios shares external dual-band antennas and the other has its own internal 5 GHz antennas.

In theory, wouldn't lower frequencies have better range? (like 2.4Ghz vs 5Ghz)

Though, it is probably fractions of a percent, if anything.

 

[pete y testing]

hi guys

i have to agree with the finding and in my own conclusions that i posted in the whirlpool forum

see

wireless AC device coverage comparison
https://forums.whirlpool.net.au/forum-replies.cfm?t=2384995

coverage has just about come to its max penetration and coverage with the existing rules regarding eirp / transmission power levels and they have done so since the r7000 was accused of going over the permitted limit

so with wireless coverage no longer being a measurable difference its other features that will make the point of difference between brands

the asus rt-ac88u has the point of difference of its 8 ports and teaming which will be wanted by those that require really high lan throughput

we are drowning in the many classes of wise ac routers available atm and i think the manufactures arnt doing them selves any favors flooding the market with release after release

im yet to fully read the post and will commend further once i have read more

pete

btw im yet to test the asus rt-ac88u or 5300 series

 

[System Error Message]

teaming is useless on consumer routers because the switch chip only has 1Gb/s links to other components (wifi, CPU, etc).

 

[pete y testing]

In theory, wouldn't lower frequencies have better range? (like 2.4Ghz vs 5Ghz)

it depends where you are located and the regulations , in australia the high 5 gig channels have a standard power rating of 4mw i think , where the lower 5 gig transmission has lower power setting according to the regulations , this i believe may be changing for usa users with the ffa changing the rules on the lower 5 gig channels but here in oz this change may never come about

 

[pete y testing]

teaming is useless on consumer routers because the switch chip only has 1Gb/s links to other components (wifi, CPU, etc).

i think you mis understand why it would be useful , say you have multiple transfers from a teamed nas or multiple streams the benefit of having twice as much supply is akin to tri band routers in that there is more supply available so each device will get more , think of it this way , 1 comp can transfer at 110MB/s from a nas to a giga ethernet connected pc , if 2 try and transfer at the same time they halve that eg 55MB/s each when used with a normal router without teaming

with teaming they share twice as much so 2 comps can both transfer at 110MB/s , so sure there are benefits to teaming a nas to a router