Linksys EA9500 Max-Stream AC5400 MU-MIMO Gigabit Router Reviewed

[sfx2000]

It's time someone starts taking a look at such shady marketing practices. It's misleading, and borderline illegal to advertise such features that clearly do not work as advertised. Not sure if the guilty party here is the SoC manufacturer (BCM) or the various router manufacturers all pushing these bullet points in their marketing material.

I would agree here - it's marketing, and the capability is there on paper, however as RMerlin points out, most folks will likely never see the full capability of the devices as they're currently being marketed.

The best improvement I've seen on the recent crop of Wave 2 devices is

• 4SS and more Radios per NIC - The impact of the 4*4:4 configs, as all clients benefit from the additional radios and MIMO capabilities - even single stream devices benefit there, and two/three stream clients even more.
• MU-MIMO - not so much for MU, but the additional compute capacity, along with enhancement client management needed for MU (sounding/ranging, scheduling, and better beamforming) benefits SU clients as well as MU clients
• DFS/TPC enhancements - was good in Wave1, so far seems even better in Wave2 - this has knock-on benefits for all bands, esp when AP's are set to auto channel, as they are now more agile in channel selection and interference avoidance/mitigation/minimizing

Things that are not so beneficial...

• TurboQAM/NitroQAM - not so much for 2.4GHz (e.g. none), TurboQAM (QAM256) in 5GHz, that's part of the Spec - but QAM1024 is definitely non-standard, and difficult to actually see due to noise floors needed - same goes with TurboQAM in 2.4GHz, most areas it's just too noisy to use...
  
• Multiple WiFi NIC's on the same band class - e.g. SmartConnect/Tri-Band - just my opinion, but the downsides (intermod RF interference in the near field), I've never really been a big fan of that configuration, but that's me...

So big numbers is one thing, but the real takeaway is that Wave 2 AP's can improve performance at a given range for any client - it's not going to be a huge step forward like AC1900 was in Wave 1, but there is improvement to be found.

 

[pete y testing]

anyone heard of this tolly group ?

they claim to have tested the ea9500 and are sprouting the following

-- Up to 57% greater individual client throughput in line-of-sight tests

-- Up to 40% greater aggregate throughput in line-of-sight tests

-- Up to 36% faster per client average client throughput in mixed distance/floor tests.

http://www.prnewswire.com/news-rele...ing-wi-fi-router-on-the-market-300310949.html

mind you the so called study was commissioned by linksys

http://tolly.com/ts/2016/Linksys/EA9500/Tolly216133LinksysEA9500MUMIMOWLANPerformance.pdf

and makes no mention of the comparative devices tested dont have working MU-mimo

 

[thiggins]

Tolly is a well-known and respected test company. All testing is paid for by their clients.
The EA9500's MU-MIMO is barely working, like all other Broadcom MU-MIMO at this point.

 

[pete y testing]

The EA9500's MU-MIMO is barely working, like all other Broadcom MU-MIMO at this point.

so why is their claims different to your conclusions and or results and who is wrong or right

i do trust the snb testing and as this so called paid for testing is so different plz explain as it makes no sense to be so different

 

[RMerlin]

The "up to" bit makes me suspicious. It implies they might be comparing a worst-case scenario versus a best-case scenario, for marketing purposes. (disclaimer: I didn't read their review, just your post).

 

[sfx2000]

Tolly is a well-known and respected test company. All testing is paid for by their clients.
The EA9500's MU-MIMO is barely working, like all other Broadcom MU-MIMO at this point.

Tolly does a good job - but consider their output as "sponsored" - e.g. one can build a test plan and report around the customer's objective - cherry picking the use cases that reflect the sponsor in a positive light.

SNB's Test Cases are more objective, and vendor neutral...

 

[pete y testing]

this my main issue with so called reviews as there is no way for the average joe to know the shyt from the sugar


so called "sponsored" reviews and testing are a blight on our landscape as the blur the facts , plus the manufactures then claim then as professional testing and evidence of how good their product is

in my environment i could not differentiate between the asus rt-ac88u and linksys ea9500 as far as wifi coverage goes

pete

p.s

why am i getting this again ?

This message is awaiting moderator approval, and is invisible to normal visitors.

 

[Kouji71]

Hi, I have a coupon for bestbuy to get this router for $200 which is a substantial discount. Is this router worth it, even at that pricepoint, or are there better options at $200?

 

[L&LD]

Hi, I have a coupon for bestbuy to get this router for $200 which is a substantial discount. Is this router worth it, even at that pricepoint, or are there better options at $200?

The RT-AC1900P would be my recommendation (along with RMerlin's firmware, of course).

If there is a further $200 discount, that would be substantial.

 

[RMerlin]

Hi, I have a coupon for bestbuy to get this router for $200 which is a substantial discount. Is this router worth it, even at that pricepoint, or are there better options at $200?

Hardware-wise this would be a very good deal. Software-wise, as long you're happy with Linksys' firmware, it'd be fine. If you need more features or want to run a third party firmware, look for another brand instead.

 

[Kouji71]

Hardware-wise this would be a very good deal. Software-wise, as long you're happy with Linksys' firmware, it'd be fine. If you need more features or want to run a third party firmware, look for another brand instead.

Yeah, I got it, but now I'm thinking of returning it.

My transfer speeds are exactly the same as the rented netgear AC1900 we had from cox, which is super disappointing. Transferring to my NAS, even when I'm in the same room as the router, maxes out at 18MBps, even though I have a ac3160 wireless card built into my laptop. I thought it was the router that was slowing things down, guess I was wrong.

but yeah, the firmware on the linksys is pretty lame. Probably going to return it just for that reason, I've been asking on the DD-WRT forums, but it doesn't look like support for this router is anywhere near complete, if it's even ever happening.

As an aside, it's funny to me that "800 Mbps wireless" gives me 18MBps transfer while 1Gbps wired gives me speeds of 120 MBps. Shouldn't they be at least a *little* bit closer? are wireless numbers really that off?

EDIT: actually, I feel stupid now ac3160 is the model number, not the speed. very (inentionally?) deceptive numbering from intel. I'll have to do some more tests with a usb 3.0 ac3100 adapter I have at home.

 

[L&LD]

I don't think there are any reasons to do further testing. That AC3160 is a 1x1 capable card (18MB/s is actually pretty respectable then). Curious what band you were using (2.4GHz or 5GHz 'AC')?

What NAS do you have? How is it connected to your network? What speeds is it capable of to another directly wired to the same router computer?

 

[sfx2000]

As an aside, it's funny to me that "800 Mbps wireless" gives me 18MBps transfer while 1Gbps wired gives me speeds of 120 MBps. Shouldn't they be at least a *little* bit closer? are wireless numbers really that off?

And this is a great example of why the marketing oriented "rating" scheme is starting to run off the rails - it is not setting the right customer expectations...

 

[Kouji71]

I don't think there are any reasons to do further testing. That AC3160 is a 1x1 capable card (18MB/s is actually pretty respectable then). Curious what band you were using (2.4GHz or 5GHz 'AC')?

What NAS do you have? How is it connected to your network? What speeds is it capable of to another directly wired to the same router computer?

I have a self-built NAS running FreeNAS (http://pcpartpicker.com/list/JzhCwV) attached to the router on a gigabit lan port. From my desktop (also gigabit wired) I get transfer speeds of 120MBps to the NAS.

The wireless card built into my laptop is connected to the 5Ghz on the router, and network properties in windows (not very accurate, I know), usually stays around 800 Mbps if I'm in the same room, down to 400-ish Mbps if I'm downstairs.

So what "ac" is my wireless capable of? or is that a pointless question?

 

[Kouji71]

And this is a great example of why the marketing oriented "rating" scheme is starting to run off the rails - it is not setting the right customer expectations...

I'm really curious about your reply, but don't know enough about the subject to truly understand it. Would you mind elaborating for me?

 

[sfx2000]

I'm really curious about your reply, but don't know enough about the subject to truly understand it. Would you mind elaborating for me?

Sure...

They pump up the numbers based on a theoretical maximum - which is never achievable in the real world..

Take this AC5400 class router... how do we make up the numbers - it's easy, we go out of spec, and put in more stuff than is needed...

2.4GHz - this is the ISM band, and 802.11n was the end of that line of development, technology wise - at the highest performance level this means 150Mb/Sec per spatial stream, which in 11n, is a maximum of 4, which equates to N600 - Broadcom (and others) have gone out of spec by implementing TurboQAM (QAM256) and Broadcom has actually done QAM1024, aka NitroQAM - which gives really big marketing numbers on the side of the box, but hopelessly not real world in any case...

5GHz - this is a bit more of a scam - as what the vendors do is put two, yes, two 5GHz Radios with 802.11ac, and again, they're now going out of spec with NitroQAM there (QAM256 is MCS9, which is in 11ac, so this is good) - the NitroQAM is non-spec, and the second radio is kind of like a third nipp1e - interesting, but hardly useful... but it does allow the OEM to advertise again, really big numbers...

Honestly though - a majority of the client side devices are much less - N150 is common in single band clients down in the cheap 2.4GHz space, and we're seeing single stream N150/AC433 client adapters out there like the AC3160 and the like... put into plain terms - that AC5400 isn't going to be any faster than N150 (in 2.4GHz) or AC433 (in 5GHz)...

The next step up, and most common in the mid-high range, is the AC867/N300 class devices - most of the higher end tablets, and many high end smart phones, along with most mid-high end laptops - this is pretty much the limit - there are exceptions - the MacBook Pro's and iMacs are most common here, that do go the the 3rd spatial stream, and there are some add-in cards, but generally, most devices are either going to be single or dual stream devices - so easentially, AC1200 class is all that is generally needed..

That being said - the AC1900 class (Asus RT-AC68U, Netgear R7000, Airport Extreme AC, Linksys WRT1900ac are examples of this class) generally do offer some value add outside of just WiFi, and that, along with the radio support is still the prime sweet spot for most consumer grade Router/AP's these days...

 

[Kouji71]

Thank you for your reply! I knew the extra antenna wasn't necessary for my house, as we only have 10 or so connected devices, but I had hoped it would provide better signal strength in other areas of the house. Doesn't seem to though, and without a custom firmware I can't change the signal power.

It's crazy, in the bedroom I get less than 2MBps transfer speeds. I think it's the giant brick chimney in between the router and the bedroom. was hoping that the router would be an easy fix.

 

[sfx2000]

It's crazy, in the bedroom I get less than 2MBps transfer speeds. I think it's the giant brick chimney in between the router and the bedroom. was hoping that the router would be an easy fix.

Can't beat physics - best performance about wifi is location of where the AP is relative to the client.

Bigger numbers are not going to punch thru that chimney...

 

[FattysGoneWild]

Pulled the trigger on this router for $264. I thought it was a solid deal. It was flash sale at Linksys and Amazon online stores. I will not miss the WRT1900AC v1 it is replacing. I have had it for over 2 years. Never again with a Marvell chipset.

 

[RogerSC]

Pulled the trigger on this router for $264. I thought it was a solid deal. It was flash sale at Linksys and Amazon online stores. I will not miss the WRT1900AC v1 it is replacing. I have had it for over 2 years. Never again with a Marvell chipset.

Like to hear about your experiences with this one. Didn't think you'd go back to Linksys *smile*.