Sign in with
Sign up | Sign in
Gigabit Wireless? Five 802.11ac Routers, Benchmarked
By ,
1. 802.11ac: The Beginning

At some point, every modern freeway was a dream to drive. The pavement was fresh and the lanes were all but empty. But inevitably, congestion set in. People learned to hop on the freeway to quicken their daily travel needs. As populations steadily grew, so did the number of cars clogging the streets. What was once a breezy late afternoon jaunt eventually became today’s four-hour exercise in asthmatic gridlock.

Automotive analogy aside, we’re really talking about the 2.4 GHz Wi-Fi spectrum. In the first days of 802.11b (circa 1999), the freeway may have posted a paltry 11 Mb/s, but there was hardly anyone else on the road. Fast forward to the present. Despite evolving through 802.11g and 802.11n, the 2.4 GHz band became a congested mess clogged with notebooks, netbooks, wireless speakers, Bluetooth peripherals, smartphones, tablets, set-tops, TVs, consoles, appliances, and all manner of other devices. These gadgets compete for what boils down to essentially three (after considering bandwidth overlap) possible transmission channels under 802.11b. A 20 MHz-wide 802.11g/n network has four such channels, while a 40 MHz-wide 802.11n network has just two.

True enough, 802.11a, which resides on the 5.0 GHz band, offers more non-overlapping channels (23, to be exact). While 802.11a offered comparable 54 Mb/s speeds to 802.11g, the 2.4 GHz solution won out in the market thanks to the fact that longer wavelengths have better ability to pass through obstructions. With an oscillation roughly twice that of 2.4 GHz, a 5.0 GHz signal was more likely to die in less distance than its competitor, and so 802.11b/g went on to become the dominant public wireless communication standard. By the time 802.11n arrived supporting both radio bands, wireless had become so popular that interference and congestion was a significant problem for many users. While 802.11n employed several technologies to help improve wireless performance, it was clear that the 2.4 GHz band was becoming ever more bogged down. For a more thorough look at these problems and some of 11n’s answers to them, we strongly recommend reviewing our two-part series, Why Your Wi-Fi Sucks and How It Can Be Helped and What To Do About It.

Today, the successor to 802.11n, dubbed 802.11ac, is far enough down the road of standardization that vendors have the confidence to start releasing product. At present, 802.11ac is working under Draft 4.0. The 802.11 Working Group expects the standard to be finally approved by late 2013, although by then the technology will already be pervasive in the market.

The first 802.11ac chipset, from Quantenna, began shipping in November 2011. In April 2012, Netgear arrived with the first 802.11ac consumer router, fueled by Broadcom innards. Other vendors soon followed. By the end of 2013, we expect mid- to high-end consumer routers to have completely switched over from 802.11n.

But for now, 11ac is new, still relatively scarce, and priced accordingly. Is it worth it? We’ve seen pre-standard Wi-Fi advances that didn’t merit their price tags in the past. Are we in for another round of disappointment, or is today’s price premium for 11ac a bargain for the performance boost? We only know of one way to find out.

2. 802.11ac Advances

Gigabit Wireless. This is the phrase that pays when it comes to marketing 802.11ac, because finally the wireless providers have a technology able to compete against structured CAT5e or CAT6 wiring. Why would you hassle with the deployment and location restrictions of wired networking if you can get the same results from Wi-Fi? You wouldn’t...if the promise held true, that is.

We saw in Gigabit Ethernet: Dude, Where's My Bandwidth? that you can get 100 MB/s+ throughput on a gigabit network using 28 feet of cable the same as 50 feet. The same story showed us it's nearly impervious to environmental interference. So, unlike wireless, we’re not left thinking, "Well, it says 1000 Mb/s, but I'm really only getting 30 MB/s." Providing you don't have any bottlenecks, gigabit is gigabit, period. As we’ll see, 802.11ac is not gigabit-class wireless. That's marketing. But is it better than 802.11n? Oh, most definitely.

To understand why 802.11ac is superior, we need to delve into some of its key advances over the prior Wi-Fi technology.

Exclusive use of the 5.0 GHz band. 802.11n makes use of either 2.4 GHz or 5.0 GHz, but we know that the 2.4 GHz range is already congested. It works, but 2.4 GHz is unreliable, and the more we want to trust it with high-bandwidth data, such as streaming HD video, the more reliability becomes a factor. Simply put, 2.4 GHz is almost tapped out, at least with current-gen approaches. You can force it into providing more performance, through “bad neighbor” tactics like channel bonding, but this has counterbalancing negative effects for others in the wireless community. The 5.0 GHz range is largely pristine land for wireless drilling, if you will, and the IEEE forces behind the new wireless standard opted to open it up for its next-gen resources.

Wider channel bandwidth. The 802.11n standard allows for the combination of two 20 MHz channels into a single 40 MHz bonded channel. In the 2.4 GHz range, already having 40 MHz channels dropped the number of effective channel options to just three. With 5.0 GHz Wi-Fi, we have 23 possible 20 MHz channels. This yields 11 effective 40 MHz channels. And now, with 802.11ac, we’re starting with 80 MHz channels, of which there are five non-overlapping options. And yes, the 802.11ac spec does scale up to 160 MHz bonded channels, but there will only be two such channels available. We’ll suspend judgment on whether 160 MHz is a good thing when we start hearing reports of how such super-wide channels perform in residential areas, particularly in the company of competing HDTV sets and smartphones.

Mo’ MIMO. Multiple-in, multiple-out (MIMO) technology provides for the separation of a single data stream into more than one sub-stream able to travel along different radio paths. This separation and recombination of signals yields higher total data throughput in many circumstances. However, more sub-signals (properly called spatial streams), results in the need for more transmit (Tx) and receive (Rx) antennas. The 450 Mb/s rates advertised in the latest, highest-end 802.11n products are only possible with a 3x3:3 (three transmit, three receive, three stream) antenna array. While 802.11n provides for up to four spatial streams, 802.11ac can use up to eight.

MU-MIMO. Multiple-user MIMO can turn multiple users into spatially disparate, but wirelessly linked transmission resources. In other words, with multiple radio terminals in a given area, all can cooperate in order to improve each terminal’s performance. The singer-user MIMO found in 802.11n can only operate with the multiple antennas hard-wired into a single terminal. With MU-MIMO, 802.11ac access points will be able to process MIMO signals from multiple clients simultaneously, rather than having to hop quickly (and inefficiently) from one to the next. This should dramatically help with airtime fairness issues in highly-populated client environments.

Optional beamforming. In Why Your Wi-Fi Sucks and How It Can Be Helped, we spent considerable time delving into beamforming and the circumstances in which it can dramatically improve wireless throughput. At the time of that writing, there were no industry standard approaches to beamforming, leaving buyers to pick among a few vendors who chose to improve their 802.11n products with proprietary approaches to the technology.

3. Broadcom: Insider Comments

As the single source for 802.11ac silicon, we figured we could interview no one more authoritative on the subject than Broadcom. We sat down with Dino Bekis, senior director of the access and wireless entertainment unit (AWE) and Richard Ybarra, technical marketing for AWE, to get their thoughts on the state of this wireless advance today and what it heralds for tomorrow.

Tom's Hardware: So now we have this new wireless spec that will push a ton of today’s current 2.4 GHz traffic onto the 5.0 GHz band. Aren’t we just going to run into the same congestion issues again a couple of years from now as everyone transitions over?

Broadcom: That possibility is always there, but with the new modulation schemes and using 20, 40, 80, and eventually 160 [MHz channels], we have more modulation schemes to work with. We have more bandwidth capability. So even though it may be congested, we still have a fatter pipe from the modulation schemes to push data. Will it get constricted? Yes, that day will eventually come, but we’re a ways away from that. Especially if we share the two spectrums we have with different types of traffic and different technology, I think it will help the overall wireless spectrum. If we share both of those bands with different technologies, it’ll help alleviate some of that congestion and allow us to do some interesting things in the 5.0 GHz space.

Tom's Hardware: Such as what?

Broadcom: Video streaming is one of our priorities. Obviously, video streaming is a major uptick as far as types of traffic across the Internet. Streaming video, downloading video, projecting video. That takes a lot of bandwidth, and that has to be bandwidth that’s dedicated. We see the 5 GHz space dealing with that pretty efficiently at this point. Maybe we send a lot of our data traffic over 2.4 and we use 5 GHz as more of a mechanism for video transfers. We’re kind of already seeing that now throughout the industry. However, I want to stop short of saying that’s a strict policy that we or any of our customers are advocating.

Tom's Hardware: Obviously, faster is better. But is there more significance to the 802.11ac transition than just another speed jump?

Broadcom: When I look at 5G [fifth-generation] Wi-Fi, four advantages stand out. First, there’s overall throughput, the ability to get really wireless gigabit Ethernet capability available to the home. Up until now, that really has not been achievable. Second, for a given throughput requirement per user and because of that higher aggregate capability, we can support more users on a given network. The radio design itself is more robust than the 802.11n design. We’re getting a lot better performance in terms of rate and range than in the past with 802.11n.

We’re also able to support much lower power for a given amount of data that needs to be transferred than with 802.11n. That’s very important for a couple of different products, whether it’s battery-powered products where you can get better life out of them or whether it’s products plugged into the wall, so they’ll be able to achieve higher efficiencies and lower power consumption, which feeds into some of the green standards for products being rolled out in the industry.

Finally, we have a much more standardized approach across the board to achieving the things than we did before with 11n. For example, we support higher throughputs, like 256-QAM, for applications such as fast synching or side-loading. In the past, these were attempted by various people in proprietary fashions—Turbo Mode, etc. That no longer is required. Before, beamforming mechanisms were proprietary, and now we have a standardized approach to beamforming that will support interoperability across various manufacturers. You don’t have these islands of technology that force you down a specific vendor’s path. You’ve got higher throughput, more users with respect to rate range, lower battery power. Those are all the key benefits to 11ac.

4. Broadcom: Insider Comments, Continued

Tom's Hardware: Broadcom has its silicon. Are we going to see other vendors share your enthusiasm for such an industry-unified approach?

Broadcom: From a market adoption perspective, we’re the only ones with a product today that’s shipping. We’ve been in production since May, and we’re seeing very strong adoption across the board. As of June, the first PCs with embedded 5G Wi-Fi were being launched, particularly from Asus at Computex. We expect that by early 2013, you’ll see some of the high-end electronics platforms, like TVs, launching with 11ac integrated. Within the first quarter, we’ll have multiple phones being launched into production with 11ac.

Tom's Hardware: Some people might still be leery of jumping on 11ac without a finished standard in place. We remember the incompatibility debacle from the days of 11g Turbo, Pre-N, Draft N, and so on.

Broadcom: Also, after the 11n battle, no one felt like a victor. So when it was time to look at the 11ac standard, every vendor made a conscious effort not to have a repeat of that mistake. Very early on, the 11ac spec gelled much more cleanly than anything that ever happened with 11n. So we’re in the final draft stages. 11ac will be ratified in Q1, and very few changes have been implemented. The different constituents converged very quickly. Where we think there will be some variances, they’ll be minor, and the expectation across the board is that they’ll be easily addressed through minor software updates. It’s not going to be driving any hardware changes. I can’t say it’s zero probability, but it’s as close to zero as anyone could claim. I think we’re very safe.

Tom's Hardware: These wider 5.0 GHz channels give us a little cause for concern. We’ve got 11ac specifying up to 160 MHz while 11n was already causing some issues at 40 MHz. Should we be worried?

Broadcom: The work has gone into making sure that 11ac doesn’t impact [5.0 GHz devices] significantly. Now, you can’t avoid the fact that as you go from 40 to 80 to 160 MHz channel bandwidth you’re gonna be using up more of the spectrum. At some point, you’re going to have a limited number of channels that you can take advantage of with those higher bandwidths. So yeah, there are ways to step back from 160 MHz to narrower channel bandwidths, but physics is going to prevent multiple 160 MHz channel bandwidth clients in the same location operating in a very high number. But the adoption we’re seeing today has 80 MHz channel bandwidth as the baseline people want to run with. 160 may apply in longer-term applications, but it’s not something we’re seeing very high demand for today.

Also, from a standards perspective, 80 MHz is mandatory and 160 MHz is not. And for what we’re doing now, 80 is sufficient. As you look at current router options, you’re going to find you’re locked into 80. People are being much more conservative in their deployment, more metered in their approach. They want to make sure there’s a successful roll-out before they take the next turn of the crank.

5. Broadcom: Insider Comments, Continued

Tom's Hardware: Is the 80 MHz limit in today’s routers enforced at the hardware level rather than in firmware? I’m sure people will want to hack their way to the higher speed if possible.

Broadcom: No, it’s not in firmware. The 80 MHz cap is currently done at the hardware level.

Tom's Hardware: With 802.11n, we had the potential for four spatial streams, and hardly anyone went past three. Now, with 11ac, we have a maximum potential for eight. Will we take advantage of it?

Broadcom: True, 11n supported up to four spatial streams. There weren’t many folks that implemented that. Broadcom did three, then two for the tablet space and one for mobile phones. Three was for high-end PCs and infrastructure. When you go to 11ac, yes, the standard supports up to eight. From what we’ve seen from our market analysis and customers, there’s an aesthetics angle. In North America and Europe, everybody loves sleek designs and embedded antennas. When you go to Asia, the more visible the antennas, the higher the perceived performance level of the device. I found that interesting from a cultural perspective. But what we’re hearing from customers is that their view in terms of trading cost and performance is three spatial streams is probably what the vast majority will want to deploy for infrastructure. And certainly, three spatial streams of 11ac over 11n will give you a 3x capability improvement. I don’t see strong interest in 11ac, at least at this point, for anything beyond that. Additional spatial streams might give you some incremental performance improvement, but you’re going to pay a significant cost premium for that.

Tom's Hardware: Look six months out and describe a common use scenario for 802.11ac and the bandwidth required behind it.

Broadcom: You know, AT&T U-verse has been a watershed launch event for this industry. You’ve seen the ads, maybe, where they have this wireless set-top box receiver out by the pool or wherever they want without having to worry about a cord being connected to it. Every one of the carriers we’ve spoken to is trying to move toward that. So you couple that with very high bandwidth, whole-home DVRs, wireless DVRs transmitting multiple programs simultaneously to different TVs. You’ve got these portable devices in the home with friends coming in or you wanting to share on a big screen. All these things take up bandwidth. So if you’re slicing and dicing today where, in the best case, you have maybe 300 to 400 Mb/s of throughput, when you’re talking about HD video streams, especially 3D, you’re anywhere from 10 to 25 Mb/s per stream. You quickly start consuming the capability of your home’s router. So moving to 11ac is going to let you share video, both at the consumer and carrier level.

6. Test Setup And Methodology

We must begin with the usual caveats about wireless testing. As we detailed in Part 2 of Why Your Wi-Fi Sucks, environmental conditions wreak havoc on residential throughput tests, such as those we’ve conducted here. However, unless one has access to an industrial-class, sealed RF test chamber or perhaps the isolation of lunar orbit, there’s nothing to do but try to pick an environment with fairly limited competing RF traffic and interference. That is, if that’s what you want. There’s a convincing counter-argument that can be made for picking a highly congested environment, as this will reflect difficult real world conditions and pose greater challenges to routers. Real world is good. Randomly fluctuating conditions are bad. Still, by looking for patterns across a diversity of tests and traffic types, we believe we can draw some fairly reliable general conclusions.

We conducted all testing in my home, a 2,650 square-foot, two-story house in a suburban area outside of Portland, Oregon. We conducted all 2.4 GHz tests using 40 MHz/auto settings on channel 1, as this channel (out of the selections 1, 6, and 11) tended to have the lowest number of visible competing access points. Similarly, we used channel 161 for all 5.0 GHz tests. Like so many other variables in Wi-Fi testing, there was some debate over this point as well. We ultimately decided on fixed channels for the sake of consistency across the routers being tested. We might have alternately selected the more popular 2.4 GHz channel 11, as higher frequencies tend to offer higher throughput, even at the risk of encountering more obstruction from surrounding traffic. Moreover, we might have left channel selection unlocked to better see how routers coped with changing environmental conditions. There is no right or wrong approach here, and we might return to test these variables in a future follow-up article.

We tested with two systems, a “server” desktop system and a “client” notebook. The server remained positioned in the home’s upper floor corner office. The client rested in either the same office at a direct line of sight distance of 10 feet from the server or it was positioned about 70 feet away, in the home’s opposite downstairs corner. In all tests, the server connected to the router via gigabit Ethernet. The client connected to a spare Netgear R6300 router in bridge mode for 2.4 GHz testing, and a Cisco Linksys WUMC710 bridge for 802.11ac work (via gigabit Ethernet). The directional orientation of the routers and bridges was kept consistent for all tests.

We conducted three basic tests. First, we created a 2.00 GB folder containing hundreds of MP3, EXE, and stray work document files. This was used to test transfer throughput speeds in both directions. Next, we turned to the networking test module in PassMark’s PerformanceTest 7 suite (we'll transition to version 8 in subsequent articles.) As a corroboration of PerformanceTest 7, as well as a deeper look at some of our traffic’s attributes, we ended with Ixia’s IxChariot. Specifically, we ran two of IxChariot’s built-in scripts. We transferred 100 records with the High-Performance TCP Throughput script and 1000 records with the UDP Throughput script.

Here are our two system configurations:

Test Server Specs
Processor
AMD FX-8150 (Zambezi) @ 3.6 GHz (18 * 200 MHz), Socket AM3+, 8 MB Shared L3, Turbo Core enabled, Power-savings enabled
Motherboard
Asus Crosshair V Formula (Socket AM3+) AMD 990FX/SB950 Chipset, BIOS 1703
Memory
G.Skill 16 GB (4 x 4 GB) DDR3-1600, F3-12800CL9Q2-32GBZL @ DDR3-1600 at 1.5 V
Storage
Patriot Wildfire 256 GB SSD
Graphics
AMD Radeon HD 7970 3 GB GDDR5
Power Supply
PC Power & Cooling Turbo-Cool 850 W
Operating System
Microsoft Windows 7 Professional (64-bit)
Test Client Specs
ModelAsus N56VM
Processor
Intel Core i7-3720QM (Ivy Bridge) @ 2.60 GHz (26 * 100 MHz), 6 MB Shared L3, Hyper-Threading enabled, Turbo Boost enabled, Power-savings enabled
Memory
Hyundai 8 GB (2 x 4 GB) PC3-12800, HMT351S6CFR8C-PB @ 1.5 V
Storage
Seagate ST9750420AS 750 GB, 7,200 RPM HDD
Graphics
Nvidia GeForce GT 630M
Power Supply
Asus ADP-120ZB
Operating System
Microsoft Windows 7 Professional (64-bit)
7. AirLive N450R And Asus RT-AC66U

AirLive N450R

AirLive is the odd man out in our round-up. The N450R is a dual-band router, but it is not 802.11ac-compatible. Rather, it uses beamforming for dual-band 802.11n. The company specifies 450 Mb/s for 5 GHz and 300 Mb/s for 2.4 GHz. On paper, this looks to be the best 5.0 GHz that one can hope to get, short of hopping to 802.11ac. Also, we wanted a low-cost, high-performance example of prior-gen technology to compare against our 802.11ac line-up. AirLive tells us the router should be priced around $116. Unfortunately, though, it isn't available in the U.S.

AirLive N450RAirLive N450R

Compared to the rich GUIs found in many top-name routers, AirLive’s menus are fairly plain, reflecting the sort of menus found in routers five or six years ago. That’s not to say they’re bad. The N450R comes stocked with all of the basic features we wanted, including thorough and understandable English documentation. As with all of the routers here, the N450R comes with four gigabit Ethernet ports and a Wi-Fi Protected Setup (WPS) button for easy connection with WPS-ready client adapters. AirLive also houses two USB ports for NAS storage and a 3G network adapter dongle.

AirLive GuiAirLive Gui

AirLive's GUIAirLive's GUI

Asus RT-AC66U

Like AirLive, Asus also employs three external antennas, but Asus goes nearly all out to maximize their potential. The RT-AC66U ($190 at Newegg) applies 3x3:3 on both radio bands, specifying 450 Mb/s for 2.4 GHz 802.11n and 1300 Mb/s for 5.0 GHz 802.11ac. To be clear, Asus does not specifically state that it uses beamforming, but the company’s "exclusive AiRadar technology" is able to "detect the direction" of connected clients and amplify their signals, which sure sounds like beamforming to us, though it could be something else. If you were so inclined, you could remove Asus’s detachable antennas and opt for higher-gain alternatives.

Asus RT-AC66U - FrontAsus RT-AC66U - Front

Our focus in this article is on hardware and performance, so we’re going to glance over a lot of usability and non-performance features. However, the RT-AC66U is a model of user-friendliness, starting with the browser-based setup process, host of internal monitoring screens, and perhaps the most attractive and intuitive menu system we’ve ever seen in a router.

Asus RT-AC66U - RearAsus RT-AC66U - Rear

Most of all, there’s AiCloud, which is a bit like having an Asus version of Pogoplug built into the router. Like AirLive, Asus builds in a pair of USB 2.0 ports, but AiCloud makes much better use of them. Essentially, any storage device (PC, NAS, USB, and so on) connected to the router can stream files through the AiCloud service and out to your Android or iOS mobile device or PC (via a Web browser). Compared to getting the same functionality through the old-school DDNS methods, AiCloud is effortless and immensely more gratifying.

Asus AiClouldAsus AiClould

8. Belkin AC1200 DB And Buffalo AC1300/N900

Belkin AC1200 DB

Bad news first. We spent about an hour on the phone with Belkin’s tech support and ended up getting a replacement unit when the first unit wouldn’t connect to our client in the long-range tests. The second unit had the same problem, which is why we won’t bother with the usual "well, maybe this particular unit was a lemon" spiel. No, the problem was that Belkin tried to undercut its 802.11ac competition ($150 at Amazon) with a 2x2 antenna design and did a terrible job of implementing it. This has "marketing decisions trumped engineering" written all over it. We don’t like bashing hardware, and we don’t have to in this case. The results depicted in our charts will speak for themselves.

In case you were curious as to why Belkin’s advertising talks about the benefits of 802.11ac rather than its own model’s specific results ("up to 2.8x faster physical data rate when compared to 802.11n Wi-Fi routers using two antennas to transmit and receive data"), now you know.

Belkin AC1200 DB - FrontBelkin AC1200 DB - FrontBelkin AC1200 DB - RearBelkin AC1200 DB - Rear

We won’t belabor the pointless. Yes, the AC1200 has four gigabit ports, some basic QoS features, parental controls, WPS push-button support, and so on. Yes, it can handle traffic on both bands simultaneously. The menu screens are nothing special. But why discuss this further? Our support rep mentioned that the company is working on a 3x3 update to this product. Wait for that if you’re interested. This 2x2 AC1200 functions at close range, but any decent 802.11n router will blow it out of the water for far less money. Weak sauce, Belkin, weak sauce.

Belkin's GUIBelkin's GUI

Buffalo AC1300/N900

On the other hand, Buffalo’s AirStation AC1300/N900 (WZR-D1800H; $160 at Newegg) holds a few pleasant surprises. While a bit on the boxy side, the relatively low price is very attractive for a model that delivers a fair amount of performance. Setup is straightforward, we like the inclusion of guest SSID access, it can serve as an access point, and it’s DLNA-certified for easy media streaming.

Buffalo AC1300/N900Buffalo AC1300/N900

For us, Buffalo’s chief weakness is its menus, which are occasionally confusing, always unattractive, and comparatively slow to refresh. We do like the explanatory text that Buffalo builds into the right side of its interface, but overall, the firmware (v1.89) is in need of a major makeover (Ed.: As of publication, the newest firmware version is 1.91, though all vendors had to compete using versions submitted by a cut-off date). If you have your doubts, check out how Asus and Linksys do things, then ask yourself which approach you'd rather take. Still, Buffalo does use a 3x3:3 antenna design.

Buffalo's GUIBuffalo's GUI

Belkin GUIBelkin GUI

9. Linksys EA6500/AC1750 And Netgear R6300

Linksys EA6566/AC1750

The Cisco Linksys EA6500 ($200 at Newegg) arrived in our hands, along with Cisco’s bundled WUMC710 802.11ac bridge ($150 separately). Based on some of our prior testing with the E3000 and other Linksys models, we had high hopes for the EA6500, and there were times when those hopes were met. However, we would recommend that, if you’re interested in the EA6500, you should read through the recent Newegg customer comments on the model. Some of the negative comments are indeed reflected in our results; others have clearly been addressed by Cisco already.

Linksys EA6500 - FrontLinksys EA6500 - Front

Linksys EA6500 - RearLinksys EA6500 - Rear

The EA6500 has a lot going for it, though. The router is simultaneous dual-band, with 3x3:3 antenna configurations in both bands. Two USB ports allow for printer sharing and add-on USB storage. DLNA support provides media serving, and QoS tools help prioritize certain traffic types.

Linksys AC1750Linksys AC1750

As with Asus, Cisco has whipped up an app/browser-based control and file-streaming platform, called Linksys Smart Wi-Fi. This gives you the ability to control several of your router’s features, such as parental control, guest access, QoS, and USB storage, straight from your handset or tablet. Additionally, there are half a dozen applications (some of which are iOS- or Android-only, while others support both) for IP security camera monitoring, network security, media streaming, and so on.

Linksys Smart Wi-FiLinksys Smart Wi-Fi

Linksys Smart Wi-Fi in ChromeLinksys Smart Wi-Fi in Chrome

As mentioned earlier, Cisco has continued Linksys’ propensity for kicking butt on its router’s built-in menus and options. While deep, the feature set is tidily swept inside of a very elegant, largely intuitive tabbed interface. That is, if you’re using a compatible browser. When we tried to log in via Chrome, all we saw was the “Log in” button and the country pull-down menu. Remember that Chrome is now the dominant PC Web browser. Oops.

Netgear R6300

Last up, we have the Netgear R6300 ($200 on Newegg). After the Asus and Linksys models, the R6300 is going to look a bit repetitive, but that’s primarily because the major router vendors have their businesses pretty much down to a pattern these days. A premium model, such as one debuting the latest Wi-Fi technology, is going to have two USB ports, simultaneous dual-band with a 3x3:3 design on both bands, a streamlined installation process, WPS, and four gigabit ports. All boxes come checked off here, and if you like Netgear’s upright, trapezoidal design, even better.

Netgear R6300Netgear R6300

Not to be outdone by Linksys Smart Wi-Fi, Netgear has its own Netgear Genie, available for Windows, OS X, Android, and iOS. Netgear Genie lets you monitor and manage your network from afar. Apple iOS users can also output to any AirPrint-compatible printer using the AirPrint app through Netgear Genie.

Because we received two R6300s, we used our second unit as a 2.4 GHz bridge, given that Netgear advertises the router as being configurable to both bridge and AP modes. However, be prepared to do some research on how to get back into the device after switching modes because it’s no longer accessible from the default IP, and Netgear doesn’t go out of its way to illustrate the process. We'll give the company fair marks for a deep, feature-rich menu system with both Basic and Advanced tab views.

Netgear GenieNetgear Genie

Netgear GenieNetgear Genie

10. Results: 2 GB Folder Copy

Our first glimpse into the performance of our six contestants reveals some interesting points. The most obvious question is: what the heck happened to Belkin? We thought this must be a fluke of some sort, but subsequent tests across this and other benchmarks corroborated our first impression. The AC1200 DB is not only crippled by its dual-antenna design, it’s not even able to perform at 802.11g levels. As mentioned previously, we got Belkin tech support on the phone, walked through all of the settings, and so forth, but there was no help for it. Belkin has an improved design in the works, and we hope to test that at some point, but for now...consider our data an object lesson in why you want your WLAN gear well-reviewed and utilizing 3x3 antenna tech. Alas, the AC1200 DB would get spanked by many 10-year-old 802.11g routers.

While we’re on the 3x3 subject, check out AirLive. By bringing beamforming onboard, the $116 N450R pulls in some remarkable results for being "just" 5 GHz 802.11n. In one direction, it even outperforms the Buffalo 11ac router. Overall, the N450R still trails every 11ac router (except Belkin), but not by much. On a performance per dollar basis, in this location and application, the N450R definitely provides a pleasant surprise, if not new legs for the current-gen tech.

Keep in mind that this is our same-room test, which theoretically should reflect optimal conditions. However, as we step back to 2.4 GHz 802.11n, transfer performance plummets. Look at the difference in uplink speed for Netgear. That's a difference greater than 600%! What causes such a drastic change? Yes, we found between four and seven competing networks in our airspace at any given time during testing, but these were fairly faint. Moreover, AirLive actually does its 2.4 GHz work with only a 2x2 antenna setup, and it still manages to trounce every competitor except Asus. That’s insane. How the likes of Linksys and Netgear, both of which were double-checked on this test, could pull in such embarrassing numbers is a mystery. Suffice it to say that we have renewed respect for Asus' product engineering and AirLive’s beamforming implementation.

As we switch to our 5.0 GHZ distance test, the tide turns. We’ve actually seen many routers over the years fail across this distance in this location. Like Belkin's solution, older hardware often won't connect at all. So, the fact that we’re measuring triple-digit Mb/s results from the four real 11ac competitors strikes us as miraculous. Also consider how little throughput loss there is between our two locations. We’re used to seeing 60% to 80% loss in these circumstances, but the 11ac routers shed almost no performance at all, and in some cases do even better at distance.

Yes, it’s great that AirLive can still pull in enough average throughput to sustain multiple HD video feeds, a feat we would not have expected, but its rivals are hitting numbers three times larger. This blows us away. This chart alone makes us want to recommend 802.11ac without reservation.

Our 2.4 GHz cross-house test doesn’t surprise. Again, Asus and AirLive dominate, Belkin can’t connect, and the other three limp by. Later on, we’ll get a better idea of what’s happening to stream integrity during these diminished transfers. Hint: It’s not pretty.

11. Results: PerformanceTest 7, Same-Room

PerformanceTest 7’s network test is similar to IxChariot in some ways, and it provides a graphically friendly way to cross-check other benchmark results. Here, in our 5.0 GHz TCP same-room test, we see Asus pull back a bit, even slipping in behind AirLive. Buffalo, Linksys, and Netgear all perform in the 165 to 180 Mb/s range, which, as a general average, meshes well with what we saw in our 2 GB transfer tests. In fact, Asus is the only player here that sees a significant difference between the two benchmarks.

In moving to UDP, throughput suddenly skyrockets, then crashes into some sort of bottleneck. Netgear's R6300 is the only router to not pass the 600 Mb/s mark.

To double-check our assumption and find out why traffic was maxing out, we went to David Wren, the creator of PerformanceTest.

“My guess at what is happening is that the device driver is just accepting an unlimited amount of data, then what can’t be sent with the available bandwidth gets thrown away,” he replied. “UDP was designed for tasks like streaming video and VoIP. So it would be like trying to push a super high definition video across the link, but on the receiving side you find that five out of every six video frames didn’t make it. But from the point of view of the sending application, everything is sent. In real life situations UDP is not used for transmitting as much data as possible as fast as possible. It is used when data needs to arrive on time, meaning that lost data isn’t worth recovering (or retransmitting) because you know that new data, e.g., the next frame of the streaming video, will arrive soon enough. You might find the CPU or PCI bus maxed out at 600 Mb/s, preventing even more data from not being transmitted.”

When we inquired as to why IxChariot’s UDP data might be so much slower than PerformanceTest 7’s (as you’ll see in a bit), Wren was quick to point out that he has never used or investigated IxChariot. However, he suggested this:

“From what I have read here, it looks like [Ixia] have implemented (by hand) their own version of TCP (with ACKs, sliding windows, and retransmissions) on top of UDP. Quote: ‘...This datagram protocol is a subset of the functionality TCP provides to ensure that data is received reliably....’ Which doesn’t really make sense to me, as no one in real life would do this. If you want a reliable connection, then you use TCP. If you want a lossy connection, then you use UDP. If I understand their paper, they are really measuring two different versions of TCP, the full Winsock implementation and their own custom coded TCP-like protocol.”

Switching to 2.4 GHz TCP, we again nose dive well below 802.11ac levels, just as in our 2 GB transfer tests. Asus easily trounces the field, with Buffalo coming in second and trailing by over 40 percent. So far, we’re having a hard time reconciling the 802.11n performance of these units, as we’ve seen prior-year models that perform better on 2.4 GHz for half the price.

In the 2.4 GHz UDP test, Netgear recovers into the 600+ Mb/s range, with overall scores running only slightly lower than in our 5 GHz pass.

12. Results: PerformanceTest 7, Across-House

Now onto the long distance TCP trials for PT7. Once again, Belkin can’t connect, which is only slightly worse than its performance in our same-room setting. Even with beamforming, AirLive takes a hefty hit on throughput, although it still holds up admirably in the 55+ Mb/s range. Asus, Buffalo, and Netgear nearly triple that number, with Asus again emerging the victor. The gap for Linksys is noticeable. Having been fans of the company’s recent consumer routers, we’re starting to wonder if Linksys might have shipped its firmware a little earlier than engineers might have liked. Hopefully, a subsequent update will help.

Not much news here with UDP. Netgear now jumps into first place, suggesting that perhaps there's 5% worth of wiggle room in our tests. If so, then all five functioning routers are statistically in a dead heat for this test.

Moving to 2.4 GHz TCP at distance, we should restate that what we see in Belkin here used to be the norm for 2.4 GHz routers at this location. All five remaining routers deserve praise for holding a connection at all in this difficult arrangement. Linksys finally shows some moxie, coming in second to Asus, the only product to break 100 Mb/s on this chart.

The UDP results are all over 600 Mb/s. Move along, nothing to see here.

13. Results: PerformanceTest 7 Graphs

One of the things we love about PT7 is its data graphs. Now that we’ve seen the raw comparative results on the previous two pages, we want to look a little deeper into the data to show throughput over time. That said, we don’t want to get exhaustive and boring, so we’re going to cherry pick our results in order to illustrate points rather than be overly redundant.

AirLive, 2.4 GHz, TCP, Location 1AirLive, 2.4 GHz, TCP, Location 1

AirLive, 2.4 GHz, TCP, Location 3AirLive, 2.4 GHz, TCP, Location 3

To begin, let’s look at the effect of distance on our AirLive router under 2.4 GHz TCP traffic. Ideally, you want to see a straight line, signaling that throughput isn’t getting hammered by interference and traffic is moving smoothly. With more distance and obstacles, the propensity to see drop-outs in the chart increases. Thanks to its beamforming, the AirLive does a very respectable job and shows minimal erratic behavior in the second chart.

AirLive, 5.0 GHz, TCP, Location 3AirLive, 5.0 GHz, TCP, Location 3

AirLive, 5.0 GHz, TCP, Location 3AirLive, 5.0 GHz, TCP, Location 3

When we switch to 5.0 GHz for the same TCP tests, we see a much different and less expected story. Our same-room test with the AirLive looks blissfully even, although we see throughput take a sudden jump about 45 seconds into testing. This might signal something like a piece of interfering equipment suddenly turning off. While we ran our tests in fairly static conditions, we still saw such plateau hopping repeatedly in our test results across vendors.

Throughput hopping aside, look at our distance test results. What looks like a fair 57.6 Mb/s on our bar chart looks more like a manic disaster here. Throughput ranges from almost 80 Mb/s down to 0 Mb/s during one precipitous drop. While a cursory glance at averages could lead one to think this router can support HD streaming at distance, you have to watch where the graph bottoms out. This is the real qualifier. If, for example, a stream needs 10 or 20 Mb/s before system overhead, than this router clearly cannot deliver reliably under these conditions.

Asus 11ac, 5.0 GHz, TCP, Location 1Asus 11ac, 5.0 GHz, TCP, Location 1

Asus 11n, 5.0 GHz, TCP, Location 1Asus 11n, 5.0 GHz, TCP, Location 1

Lest anyone think we’re only picking on AirLive, let’s examine a TCP quartet from Asus. Looking at 11ac same-room performance, we get a little jiggle in the opening second or so as the connection stabilizes, then a long, stable plateau at 90+ Mb/s, then a sudden jump into the 140+ Mb/s range. Then when we switch to 802.11n, all semblance of stability vanishes. Performance swings across a 100% range from 70 to 140 Mb/s. This is obviously workable from an application standpoint, but it shows the striking variability of 11n throughput, even for such an excellent-performing router.

Asus 11ac, 5.0 GHz, TCP, Location 3Asus 11ac, 5.0 GHz, TCP, Location 3

Asus 11n, 5.0 GHz, TCP, Location 3Asus 11n, 5.0 GHz, TCP, Location 3

Going back to 11ac in our distance setting, we once more see a fleeting ramp-up in the connection, then an impressively narrow throughput band around 145 Mb/s. Seeing this performance level across the house still makes us giddy. Our 802.11n distance test shows another throughput leap in mid-test, but we again see a much broader performance band within the main plateau. Note, however, that we’re not seeing those drastic pits with Asus. Once it locks onto a throughput level, it’s very good about maintaining a floor.

Belkin 11ac, 5.0 GHz, TCP, Location 1Belkin 11ac, 5.0 GHz, TCP, Location 1

Bufflo 11ac, 5.0 GHz, TCP, Location 1Bufflo 11ac, 5.0 GHz, TCP, Location 1

Linksys 11ac, 5.0 GHz, TCP, Location 1Linksys 11ac, 5.0 GHz, TCP, Location 1

Netgear 11ac, 5.0 GHz, TCP, Location 1Netgear 11ac, 5.0 GHz, TCP, Location 1

Finally, let’s look at our remaining four routers under best-case TCP conditions. Even without looking at the y-axis numbers, Belkin is obviously out. Buffalo has the most stable, perfect-looking chart of the bunch, though Netgear gives Buffalo an interesting challenge. While it has that little warm-up blip, Netgear’s sustained throughput is slightly better than Buffalo’s. Linksys looks much more erratic, but watch those y-axis values. Numbers over 300 Mb/s for TCP? Dang!

Why we sometimes see those performance plateau shifts remains a mystery. Since we’re only seeing upward shifts in our data, in future testing we may want to look at patterns over longer periods of time, say a half-hour or more. Maybe these stable plateaus aren’t so stable with a longer time scale. These shifts happen in both test locations, so it’s not a local effect, nor does there seem to be a correlation with router/bridge combinations. It could have something to do with the TCP/IP stack, but digging into this will require additional research. For now, it remains a question mark to revisit another day.

14. Results: IxChariot, Same-Room, 5.0 GHz

Last up, we have Ixia’s IxChariot, which is probably the most widely used and trusted benchmark in wireless networking.

In our same-room 802.11ac testing for TCP traffic, Belkin sells the only router that falls flat. Even the next-lowest performer, Linksys, averages 160 Mb/s, which is stunning for TCP. Note that AirLive continues to impress at 189 Mb/s, solely on the basis of its beamforming. Just imagine when 11ac vendors cycle through their first wave of routers and decide to implement optional beamforming in their second-gen follow-ups in 2013 or 2014! Now, remember to approach those huge top-end numbers for Asus, Buffalo, and Netgear with caution. For example, here’s what was going on under the hood in Asus' IxChariot throughput graph:

Asus 11ac, hi perfTCP, Throughput, Location 1Asus 11ac, hi perfTCP, Throughput, Location 1

See the plateau shift again? For sure, if we could count on Asus to average 320 Mb/s consistently, we’d bow down and worship this router. However, until we get a better fix on why these shifts keep happening, we’ll keep our adoration in check.

Netgear 11ac, hi perfTCP, Throughput, Location 1Netgear 11ac, hi perfTCP, Throughput, Location 1

For Netgear, the situation is very similar. While its plateau averages are slightly higher than Asus', and its peak-to-trough range is similar, Netgear’s plateau shift arriving later in the test skews the average throughput downward.

Linksys 11ac, hi perfTCP, Throughput, Location 1Linksys 11ac, hi perfTCP, Throughput, Location 1

We did not break out response numbers in these analyses since they were essentially inverse mirrors of our throughput charts. Take Linksys’s response graph (above) as an example. One can say that 0.5 second is the average response time, but this clearly depends on when in the test cycle we’re talking about.

If that’s not confusing enough, let’s switch to UDP traffic for the same-room set. Remember how we redlined most UDP results in PerformanceTest 7? IxChariot’s UDP script clearly limits bandwidth—so much so that our UDP numbers underperform our TCP results, which almost never happens.

Yes, our UDP throughput is roughly half of what we saw with TCP. However, regardless of the ways in which IxChariot constricts or relaxes data flow through its scripts, we still have an accurate relative ranking of routers all adhering to the same test process. And within these rules, we see all four of our serious 11ac routers averaging a statistical dead heat.

The plateaus we see in our TCP tests don’t appear here. On the other hand, we see a startling difference in throughput patterns. Consider these two graphs from Buffalo and Linksys. The averages are very close, but the patterns obviously differ.

Bufflo 11n, UDP, Throughput, Location 1Bufflo 11n, UDP, Throughput, Location 1

Linksys 11ac, UDP, Throughput, Location 1Linksys 11ac, UDP, Throughput, Location 1

Which graph is better? We would argue for Linksys. Whereas Buffalo seems to be struggling to keep pressed against a 118 Mb/s ceiling, Linksys has a very defined floor of around 112 Mb/s. When it comes to maintaining quality of service in a stream, the latter pattern is clearly preferable.

15. Results: IxChariot

Crossing over to 2.4 GHz 802.11n in the same-room tests, we have more weirdness. For once, Belkin does not have the lowest number on the chart. Whereas AirLive, Asus, and Buffalo are do a respectable job of not letting throughput drop under 50 Mb/s, Linksys and Netgear both have instances where they bottom out at 5 Mb/s. Even Belkin manages a fractional improvement over that. Sure, Belkin has the worst average, but we’re looking under every rock for good news at this point. Because we’re working with 802.11n, we should be dealing with a level playing field stocked with vendors handling a fully mature and refined technology. So it’s interesting to see AirLive edge past a respected name like Buffalo and fully trounce Linksys and Netgear. Only Asus manages to keep the small upstart in second place (by a gaping margin).

With our UDP chart, we see the throughput numbers shift relative to TCP. AirLive and Asus nudge down a bit while Buffalo, Linksys, and Netgear all edge up. Since Linksys stands out as the chart’s highest outlier, let’s take a look at that.

Linksys 11n, 2.4 GHz, UDP, Throughput, Location 1Linksys 11n, 2.4 GHz, UDP, Throughput, Location 1

Pretty telling, right? The top spike on this chart is almost five times higher than the clearly visible average line. Just as disturbing are the many dips into the sub-20 Mb/s range. Keep in mind that this is a same-room test. We are left acknowledging that there is a fair amount of ambient noise and unpredictability in our test environment, but, at the same time, it’s a fairly average suburban scenario, and this is what routers need to cope with in the real world. We find the fact that Linksys and Netgear struggle here disturbing.

Now for the cross-house 11ac test with TCP traffic. Again, we see Belkin failing to connect, and AirLive finally manages to walk off a cliff. In fact, here’s what life looks like at the bottom of that cliff:

AirLive, 5.0 GHz, hi perfTCP, Location 3AirLive, 5.0 GHz, hi perfTCP, Location 3

The good news is that AirLive managed to transmit all 100 IxChariot test records. The bad news is that most of those records came in two bursts, like flashbulbs in the dark, and the rest of the time saw almost no throughput. AirLive aside, we’re very impressed with the TCP results for our remaining competitors, although Linksys does noticeably lag behind the other three. With an average of roughly 180 Mb/s for Asus, Buffalo, and Netgear, this compares very well with the approximately 240 Mb/s averaged by those three in our similar close-distance test. A 25% throughput loss under such difficult conditions is actually phenomenal.

UDP across the house 11ac is definitely slower under 802.11ac, but still very usable and reliable in most cases. Netgear now turns in a stable performance with the best minimum throughput rate on the chart. Asus wins on the average number, but look at the deeper test chart:

Asus 11ac, 2.4 GHz, UDP, Throughput, Location 3Asus 11ac, 2.4 GHz, UDP, Throughput, Location 3

For a long-distance test through flooring and walls, Asus’s stability here is outstanding. We only see one major blip, and we’re guessing that some random bit of ambient interference clobbered throughput for an instant, and the router responded by boosting power to compensate. When positive ambient conditions returned, the router dropped power back to normal levels. That’s a guess, but no matter what, this chart illustrates the Asus product’s ability to hold an 802.11ac signal with excellent stability and respond very quickly to adverse conditions.

16. Results: IxChariot, Across-House, 2.4 GHz

We’ll move quickly through the 2.4 GHz cross-house tests as these results largely repeat what we’ve observed previously.

Again, AirLive and Asus prove their strength in TCP testing through their minimum throughput times. AirLive in particular shows a narrow delta between minimum and maximum results, which is good. Of all the routers here, Asus emerges as the easy victor on this test, with Linksys coming in a distant second.

With UDP traffic, the story repeats. Asus is the only router we’d trust with a high-def stream, although we’d settle for the Buffalo in a pinch.

Asus 11n, hi perfTCP, Throughput, Location 3Asus 11n, hi perfTCP, Throughput, Location 3

Asus 11n, UDP, Throughput, Location 3Asus 11n, UDP, Throughput, Location 3

The difference in data patterns between our two traffic types can be striking. With 1,000 records passed using UDP, compared to 100 for TCP, one gets a better sense of where the “normal” bandwidth of UDP rests. In contrast, TCP seems much more meandering and variable.

One cool feature within IxChariot is its automatic report of how many bytes of data are lost in a given UDP communication. This is something you generally don’t see reflected in speed results. Does it matter if you see 200 Mb/s performance if only half of your data shows up? Maybe, depending on the application. This is why we took some of our long-distance data and broke out the bytes lost into separate charts.

Belkin scores 100% loss because it wouldn’t connect at this range. While not technically accurate since there was no actual data transmission, we felt it important to visually represent a worst-case.

The difference between 5.0 and 2.4 GHz is remarkable. Now we see AirLive’s adherence to prior-gen 11a, even with the assistance of beamforming, start to become a serious obstacle. Buffalo and Linksys have zero loss under 11ac, which is phenomenal.

Our situation reverses with 2.4 GHz. Buffalo and Netgear drop over half of their packets, and AirLive isn’t far behind. Only Asus manages nearly flawless reliability. This should be of particular note to users who anticipate using clients on both radio bands.

17. 802.11ac: A Substantial Step Up From 802.11n

Again, we’re going to ignore Belkin here. Perhaps a firmware update will bring the AC1200 back from oblivion, but we’re not going to hold our breath.

In evaluating Buffalo, Linksys, and Netgear, one can cherry pick results to arrive at a favorite. To our eyes, there really isn’t a clear winner in our charts. Linksys and Netgear definitely have an edge in the menu interface department, and we like Linksys in particular for its richer Smart Wi-Fi app platform.

If you’re on a budget and aren’t close to adding clients with 802.11ac support, AirLive remains a surprisingly compelling option. You won’t find bells and whistles, but you will get some of the best performance we’ve seen in a mid-range router. It's unfortunate that the company doesn't have anything for sale in the U.S.

Then there’s Asus, which won this contest hardly breaking a sweat. The company simply out-engineered its competition and came up with a best-in-class feature set. Even more amazing is that the RT-AC66U accomplished this in essentially a first-generation product and delivered it at a price point that handily bests its closest competition. For all of the above, we believe the RT-AC66U deserves our infrequently-bestowed Elite award.

Stepping back, are we ready to give our blessing to 802.11ac and recommend that you run out and start investing in it? Yes. Obviously, many vendors still have work to do. We want to circle back later and investigate factors like the impact of channel selection on performance, maximum usable range for 802.11ac, and other variables we deliberately tried to isolate. There's also the issue of maximum throughput, since we saw numbers that might have indicated the limits of our test systems' storage subsystems. For now, though, we’ve seen enough to believe that 5G Wi-Fi is ready for primetime.

We’d hoped to see real-world sustained transfer rates in excess of 300 (or at least 200) Mb/s. That didn’t happen. Perhaps it will take beamforming, more antennas, and other enhancements to get us there later in 2013. But we can live with 150-ish Mb/s in the same room if we’re also seeing 100 to 150 Mb/s transfer rates across significant distances through multiple barriers. For us, that was absolutely huge. When done right, 802.11ac doubled what we could achieve with 802.11n. That alone is worth buying.

Folks, the time to start your 802.11ac adoption is now.