Linksys WRT1900AC AC1900 Dual Band Wireless Router Review

[thiggins]

Review updated: Storage retest

I identified the problem with my standard USB dock. So retested with it and added eSATA results.

 

[hewligun]

Tim if it were you and looking into your crystal ball.. Between the 68u, r7000
And this linksys which would you personally pick. Price not a factor. . Assuming firmware updates improve all of them both mAnufacture and 3rd party.

 

[thiggins]

Tim if it were you and looking into your crystal ball.. Between the 68u, r7000
And this linksys which would you personally pick. Price not a factor. . Assuming firmware updates improve all of them both mAnufacture and 3rd party.

I am not a good test case. I have yet to find a router that doesn't meet my simple needs. (Only two people in household, Ethernet where I need it.)

There is no perfect product. The question is what kind of pain you like. It is early days for the WRT1900AC. ASUS has certainly had plenty of time to sort out its problems, but they still persist. The R7000 has been my suggestion and if you must buy now and must buy AC1900, would still be.

 

[ehume]

So, you and the Tweaktown reviewer report on the throughput in the two router bands. Everybody does. But I have not seen anyone report specifically on network protocols.

What I want to know is not how it did on the 2.4GHz band, or the 5GHz band, but how it did on 802.11n. That protocol uses both bands. So how did this router do on the -n? How did the R7000 do?

I returned an ASUS RT-AC68U because, despite giving my ASUS PCE-AC68 great 1900AC throughput into the basement, it could not deliver 802.11n to my kids' iPod and laptop. So I sent it back and returned my 2009 WNDR 3700 back to service. That gives far better 802.11n.

So, how to the R7000 and the WRT1900AC compare on 802.11n? How do they compare to old routers? Are they worth getting if you must support legacy devices?

 

[thiggins]

The How We Test article has the information you want. Test client is an ASUS PCE-AC66, 3x3 AC1750 class client. So 2.4 GHz band results are with an N450 class client. 5 GHz is with AC1300.

We do not test interoperability with older devices. There are just too many alternatives to make a comparison using one or two devices worthwhile.

AC is still in early days and compatibility / interoperability problems are still a work in progress. The industry pushes toward higher throughput and complexity (MU-MIMO, 4x4 and higher order chipsets), not stability.

 

[klas]

Not much performance gain, expensive, fan... What's the point? R7000 with dd-wrt, cheaper and faster.

 

[hewligun]

Tim in the other thread you mentioned you don't think this router will have much in gains as far as wireless goes. Is that because the speeds already seem consistent? Is the Broadcom chipset just that much better? You don't think linksys can release a firmware that would make it faster.

I see what you mean. They all would
Meet my needs but I buy a router once my old one does and i buy the best at the time. Seems like you think I should keep my r7000. Kind of disappointed to hear that.

 

[sm00thpapa]

First off thank you Tim for this review.

I was very depressed after reading your review. First why this has a Marvell CPU and not a Broadcom CPU is beyond me. No 256-QAM on N only and no AC only. I have a separate WRT54G V4 for my G clients. I like to leave all my N clients on one router so I always set my main router to N only. But with no 256-QAM guess that will have to change and I am not happy about it. My main reason for purchasing a router is for the wi-fi range. I have 23 devices connected at all times and 20 of them are wireless. 3 on the 5 GHz band and 17 on the 2.4 GHz band. I have 1 dead spot room upstairs on the opposite side of the house to where the router is downstairs. I am hoping that this router cures that issue. But in your review it seems it won't. Once I receive this router it will replace my tried and trusty RT-N66W for evaluation. I will post my findings once I begin using it. Thanks again for the review Tim.

 

[jalyst]

FYI the openWRT side of this router is apparently DOA & likley will never happen: https://lists.openwrt.org/pipermail/openwrt-devel/2014-April/024589.html, e.g. don't hold your breath for *any* third-party firmwares.

Yes I saw that referenced by Lom in a dramatic sort of way over at dd-wrt forum...
I don't see how it's unequivocal evidence that they're not working with OpenWRT, or doing any sort of F/OSS.

If you click on the 'next message' link you'll start to see openWRT devs response to Belkin's attempt to submit patches to the code base... which are basically "You can't do that" and "GTFO"... This thread will help you read between the lines:
http://www.dd-wrt.com/phpBB2/viewtopic.php?t=256298&start=45

This sorta thing is sooo common in collaboration between Co's & F/OSS communities that it's not even funny any more.
It could devolve into a longer-term systemic issue on Belkin's part, or, it could be a "storm in a teacup" (which is so often the case), we just don't know yet.

 

[speedingcheetah]

mmmm....PC World says its the fastest router they have tested....a bit contradictory to the review posted here in comparison to other models.

http://www.pcworld.com/article/2143623/linksys-wrt1900ac-wi-fi-router-review-faster-than-anything-we-ve-tested.html

Their 2.4ghz bad tests greatly favored the Asus 68U, but that is because the Linksys stayed in 20mhz mode...as he mentions.

I am not sure what sites to trust these days...I have heard that some of them get paid to say good things about any product...or they way they test them aren't as real world as they should be.

 

[thiggins]

Papa,
No need to be depressed. Jeez, it's just a router!

Vendors are always exploring new technology / chipsets. No one wants to be single-sourced for key technology, which they have been with Broadcom for AC. Qualcomm's first gen AC didn't cut it, so Belkin looked at Marvell.

Belkin needed to make a big splash and have something unique for marketing purposes to put Linksys back on the map. So Marvell's three-stream but four antenna solution fit the bill. More is always better in wireless and physologically, four antennas must be better than three, right?

Why Marvell agreed to get back into the consumer market after being out, is a puzzle, but they did.

Marvell's technology seems equal to Broadcom's in chamber testing. The real question is how it does in real world, which only real-world testing will answer.

If you have a range problem, add an AP or two. Informed consumers like many of you in the forums should know better than to drink the marketing Kool-Aid of one router doing whole-house coverage. The problem set is just too large for this. Wireless is not a one-size-fits-all technology.

 

[thiggins]

Regarding the WAN-LAN throughput. I'd like to see a benchmark once QOS is turned on as far as I know HW acceleration does not work once QOS is in use.

There must be a reason why the throughput is much worse on the Linksys, maybe it is because they are not using some sort of HW NAT acceleration.

It would be cool if you could try that and see if this makes a difference.

Tried a quick test enabling Media Prioritization, putting the one computer into the High Priority bucket and setting download bandwidth to 1,000,000 Kbps.

No effect on downlink throughput. Uplink was reduced ~100 Mbps.

 

[Shikami]

I do know, for those that are disappointed with features of the firmware, that they are going to be adding OpenVPN to the router. There are a few things that will come later with this router. Its not bad, but it is not a show stopper.

Although, the true test will be with OpenWRT and how feature full and speedy (or not) this router can be with it. I think, and hope Tim will re-test when OpenWRT is available.

 

[thiggins]

I do know, for those that are disappointed with features of the firmware, that they are going to be adding OpenVPN to the router. There are a few things that will come later with this router. Its not bad, but it is not a show stopper.

Although, the true test will be with OpenWRT and how feature full and speedy (or not) this router can be with it. I think, and hope Tim will re-test when OpenWRT is available.

I haven't been retesting other current routers with alternative firmware. Previous tests have always shown no performance improvement via alternative firmware. Is your experience different?

 

[speedingcheetah]

I haven't been retesting other current routers with alternative firmware. Previous tests have always shown no performance improvement via alternative firmware. Is your experience different?

That would be a bit of an undertaking. Some router models have quite a few different flavors and sub-builds of alternative firmware available.

In my experience, switching from stock firmware to alternative firmware has always been good. I went from stock Netgear to DD-WRT on a wnd3000(?) (though that router died not to long after, as every Netgear product I have owned has)....Tomato(Toastman builds) on my old(though many still in service) Cisco E2000/3000, worlds of difference in performance from stock. Going from stock Asus firmware on my old N56U to Padavan firmware, noticeable feature and performance jump.

 

[thiggins]

Can you elaborate on the specific performance improvements? Wireless range? Throughput? Stability? Same for wired routing?

 

[hewligun]

mmmm....PC World says its the fastest router they have tested....a bit contradictory to the review posted here in comparison to other models.

http://www.pcworld.com/article/2143623/linksys-wrt1900ac-wi-fi-router-review-faster-than-anything-we-ve-tested.html

Their 2.4ghz bad tests greatly favored the Asus 68U, but that is because the Linksys stayed in 20mhz mode...as he mentions.

I am not sure what sites to trust these days...I have heard that some of them get paid to say good things about any product...or they way they test them aren't as real world as they should be.

Tweaktowns review seems more in tow with what pcworld said. I am not discrediting SNB. But perhaps he had a device on the 5ghz network that he didnt know that was older and slowing it down. there is a lot of factors.

I am waiting for a few more reviews to see what trend unfolds. I have til sunday to either swap or keep my r7000. It has been fine the past 3 days so maybe the problems fixed themselves on it.

 

[thiggins]

Tweaktowns review seems more in tow with what pcworld said. I am not discrediting SNB. But perhaps he had a device on the 5ghz network that he didnt know that was older and slowing it down. there is a lot of factors.

He who, me? There was only one client device, the ASUS PCE-AC66, connected for sure since both router and client are in sealed RF chambers.

PCWorld's review specs jperf used for testing. Jperf can produce higher than real-world results depending on the settings used.

TweakTown's review doesnt' specify the application used for testing.

 

[hewligun]

He who, me? There was only one client device, the ASUS PCE-AC66, connected for sure since both router and client are in sealed RF chambers.

PCWorld's review specs jperf used for testing. Jperf can produce higher than real-world results depending on the settings used.

TweakTown's review doesnt' specify the application used for testing.

Oh i know. but he compared it against the asus which he used the same testing. I was looking more at the two head to head vs the numbers themselves. Although I did peak at the numbers.

good to know you only used one device. if that was stated in the review I overlooked that part.

I am glad to see both sides of reviews to be honest. Good to see how different people see it. I originally bought the r7000 based off SNB when most other sites said the asus was the way to go. I tried the asus as well though.

No 2 setups are the same and that could be why 1 router worked better for you and one worked better for them. It seems asus/netgear and linksys all have a pony in the race though.

 

[speedingcheetah]

Can you elaborate on the specific performance improvements? Wireless range? Throughput? Stability? Same for wired routing?

All of the above. Its been a while and I did not think to make a written record of it at the time. ( I may have mentioned some of the numbers in previous posts way back when I first joined SMB or somewhere else on the web) Only particular details I remember and was mostly concerned about was faster file transfer speed both over LAN and wifi...that was/is my main test. Tomato and Padavan firmware beat stock each time.

I did try Merlin firmware on my AC56U awhile back, but my hardware doesn't seem to like it....I can't explain why, stock just runs smoother and gives me better performance in this particular case.