|At a glance|
|Product||ASUS Dual Band AC2900 Wireless Router (RT-AC86U) [Website]|
|Summary||Three-stream 2,4 GHz and four stream 5 GHz AC class router with gigabit Ethernet ports, USB storage sharing and MU-MIMO support.|
|Pros||• Four stream 5 GHz radio|
• Full complement of ASUSWRT features
• Strong wired routing performance
|Cons||• Auto 20/40 in 2.4 GHz could cause connection drops|
Typical Price: $139 Buy From Amazon
Updated 11/16/17 - Added Storage Performance results
ASUS' RT-AC68 has had a long run. It is one of the company's best-selling wireless routers, available in multiple flavors to appease consumer electronics retailers who want to confuse consumers the same way mattress stores traditionally have.
In the quest to keep cash registers ringing, ASUS is following a similar route taken by NETGEAR with its R7000P, by refreshing the design with Broadcom's newest AC radios and a faster CPU. But to bump its "class" number higher, ASUS has given the 86U a 4 stream 5 GHz radio instead of matching the three streams on the 2.4 GHz side.
Like the AC68, the AC86 will also sell under the aliases shown in the FCC document excerpt below.
ASUS RT-AC86 versions
ASUS has stuck with the stand-up only design shared by the AC68 series, but swapped the diamond-pattern front styling for a cleaner angular style with red accents. The callouts below show a similar set of front panel LEDs and rear panel connectors. The right side also still sports WPS and Wi-Fi On/Off buttons.
Never one to be outdone, ASUS is one-upping NETGEAR by seeing NETGEAR R7000P's change from AC1900 to AC2300 and raising an extra 600 Mbps to AC2900. The path to AC2900 is via the use of Broadcom's newer BCM4365E 3x3 radio SoC performing 2.4 GHz duties and its BCM4366E 4x4 radio SoC holding down the 5 GHz side. Since both support Broadcom's non-standard 1024 QAM, the inflated "class math" comes out to 750 Mbps for 2.4 GHz + 2167 Mbps for 5 GHz = 2917 Mbps, which ASUS has shockingly rounded down to 2900.
ASUS has made a smart move by increasing the number of 5 GHz streams, which should help improve performance in that band. But precious few devices—ASUS' PCE-AC88 PCIe desktop adapter being the most notable—support 1024 QAM and you need very strong signals to reach the advertised maximum link rates.
So for most of us, the AC86 should be considered AC2300 class with maximum link rates of 600 Mbps in 2.4 GHz and 1733 Mbps in 5 GHz with devices that support 256 QAM. But don't confuse this AC2300 with the NETGEAR R7000P's inflated AC2300, which is really AC1900. Can you see why I've stopped ranking products by wireless "class"?!
Here's the RT-AC86U board with heatsinks and RF can tops removed. It's a pretty clean design, using RF front ends vs. power amps and their associated switches and LNAs.
ASUS RT-AC86U board top
Here's the top view of the original RT-AC68U board. A glance at the left-side area above the two radio SoCs illustrates my point about the AC86's cleaner RF front end.
ASUS RT-AC68U board top
If you were paying attention, you might have noticed an extra little white rectangle near the center antenna on the topside photo. That's the RF front end for the fourth 5 GHz transmit/receive chain. The photo below shows where the fourth antenna is hiding.
ASUS RT-AC86U inside with fourth antenna detail
Here's the component summary with the AC68U and NETGEAR R7000P included for comparison.
|ASUS RT-AC86U||ASUS RT-AC68U||NETGEAR R7000P|
dual core ARM v8 Cortex A53 @ 1.8 GHz
dual-core ARM Cortex A9 @ 800 MHz
dual-core ARM Cortex A9 @ 1 GHz
|Switch||In BCM4906||In BCM4708A||In BCM4708C0|
|RAM||512 MB||256 MB||256 MB|
|Flash||256 MB||128 MB||128 MB|
|2.4 GHz Radio||- Broadcom BCM4365E 3x3 11abgn-ac SoC
- 2.4 GHz RF front end (x3)
|- Broadcom BCM4360KMLG
- Richwave RTC6649E 2.4 GHz Power Amp (x3)
| - BCM4360KMLG 3x3 11abgnac SoC
- Skyworks SKY85319-11 2.4 GHz front end (x3)
|5 GHz radio||- Broadcom BCM4366E 4x4 11abgn-ac SoC
- 5 GHz RF front end (x4)
|- Broadcom BCM4360
- SiGE 5023L 5 GHz Power Amp (x3)
|- BCM4365E 3x3 11abgnac SoC w/ MU-MIMO
- QPF4519 Quorvo 5 GHz front end (x3)
Table 1: Component summary
I'm not going to dive into feature details in this review, as ASUSWRT's features are well-known. I'll be doing only component analysis and performance review. But so you're not confused about the feature set you're buying with the AC86, here are a few screenshots to show key feature sets.
Here's the Game Boost menu expanded, with everything you get with ASUS' "Gamer" routers.
Game Boost menu
The AiProtection feature set also looks complete.
Finally the Adaptive QoS screen also looks pretty familiar.
Adaptive QoS menu
I had no luck getting the AC86 to recognize either my standard Startech USB 3.0 eSATA to SATA Hard Drive Docking Station (SATDOCKU3SEF) or a WD My Passport Ultra drive after multiple attach/detach cycles and router reboots. So, no storage performance results.
After some encouragement from SNBForums regulars, I decided to give storage testing one more shot. Maybe it was the router reboot, but this time I had no problems mounting the standard drive plugged into the USB 2.0 and 3.0 ports and ran the full set of tests. All the data is now in the Router Charts; I'll be comparing USB 3.0/NTFS results only.
The AC86 did pretty well, beating NETGEAR's R7800 for both write and read with 71 MB/s and 104 MB/s, respectively. Of course, I had to change the Reducing USB 3.0 Interference setting on the 2.4 GHz Wireless - Professional page from its default enable setting to disable to achieve these results.
USB 3.0/NTFS storage performance comparison
The AC86 was loaded with 184.108.40.206.382_18219-g76de09e firmware and tested with our Revision 10 router test process. I didn't retest the RT-AC68U with the Revision 10 process. So I've included results for NETGEAR's R7000P instead, since I'm considering it the AC86U's peer.
|Test Description||ASUS RT-AC86U||NETGEAR R7000P|
|WAN - LAN Throughput (Mbps)||938||941|
|LAN - WAN Throughput (Mbps)||941||940|
|HTTP Score - WAN to LAN (%)||57.1||0.3|
|HTTP Score - LAN to WAN (%)||57.3||22.5|
|Bufferbloat Score- Down Avg.||575||513|
|Bufferbloat Score- Down Max.||446||376|
|Bufferbloat Score- Up Avg.||529||423|
|Bufferbloat Score- Up Max.||388||322|
|CTF Score (%)||99.1||32.6|
Table 2: Routing throughput
The WAN - LAN and LAN - WAN throughput benchmarks are the least meaningful, since most products can hit these numbers due to the common use of Cut Through Forwarding. The HTTP and CTF Scores are now where you want to focus.
HTTP Score comparison - WAN to LAN
Plot key file size: [A] 2 KB, [B] 10 KB, [C] 108 KB and [D] 759 KB file
The HTTP Score plots show the AC86U spanking the R7000P for both download and upload, reaching 100% of baseline throughput for both 108 KB and 759 KB file sizes.
HTTP Score comparison - LAN to WAN
Plot key file size: [A] 2 KB, [B] 10 KB, [C] 108 KB and [D] 759 KB file
Comparing average Bufferbloat scores from Table 2 above, both routers did pretty well. Converting the AC86U's scores back to latency ((1/score) x 1000) yields 1.7 msec average WAN to LAN and 1.9 msec LAN to WAN.
The Cut Through Forwarding tests look for throughput reduction when various router features are used. The 99.1% score means testing did not show appreciable throughput reduction for any of the features tested (adaptive QoS, Bandwidth monitor, Traffic Monitor, AiProtection w/ all features enabled, Parental Control w/ all features enabled and keyword blocking).
In all, the AC86U's routing looks pretty fast, no matter which features you use.
The RT-AC86U is Wi-Fi Certified. It was loaded with 220.127.116.11.382_18219-g76de09e firmware and tested with the Revision 10 wireless test process . The router was reset to factory default, then set to Channel 6 for 2.4 GHz and Channel 40 and 20/40/80 MHz bandwidth for 5 GHz. WPA2/AES encryption was used for all connections.
The Revision 10 process uses 20 MHz bandwidth for 2.4 GHz tests for throughput vs. attenuation, but uses 40 MHz for peak throughput tests. These settings are enforced by the octoScope Pal test client. 2.4 GHz channel bandwidth was initially set to 20/40 MHz, but the octoScope Pal test client kept disconnecting. Reliable connection was maintained when 20 MHz bandwidth was chosen for throughput vs. attenuation testing and 40 MHz bandwidth for peak throughput testing.
The router body was centered on the test chamber turntable with all antennas vertical as shown in the photo below. The 0° position for the router had the front facing the chamber antennas. You might think the 86U's antennas are too high for the chamber antennas to properly capture their signal.
ASUS RT-AC86U in test chamber
But I also tried the position shown below and found the standup position yielded better results, especially on 5 GHz. This is likely due to the vertical fixed position of the internal fourth 5 GHz patch antenna.
ASUS RT-AC86U in test chamber - face down
We'll start by comparing average throughput to get an initial feel for how the AC86 measures up. Average 2.4 GHz throughput of 64 Mbps downlink and 62 Mbps uplink slotted the AC86U in second and fourth place in the charts.
2.4 GHz average throughput comparison
5 GHz average throughput of 164 Mbps downlink and 263 Mbps up, landed the 86U in sixth and third positions.
5 GHz average throughput comparison
For the throughput vs. attenuation plots, I'm comparing the AC86 with NETGEAR's R7800 and R7000P and ASUS' GT-AC5300. For 2.4 GHz downlink throughput, only the R7000P breaks from the pack with much lower throughput under stronger signal (lower attenuation) conditions.
2.4 GHz Downlink Throughput vs. Attenuation
For 2.4 GHz uplink, the NETGEAR R7800 is the outlier, but in a good way.
2.4 GHz Uplink Throughput vs. Attenuation
5 GHz downlink is, as usual, more interesting, showing the greatest performance difference. Even though the AC86U runs four streams in 5 GHz, it acts more like the three-stream NETGEAR R7000P. The two four stream routers with all external antennas of the same design both perform better.
5 GHz Downlink Throughput vs. Attenuation
The 5 GHz uplink plot shows a tighter range of performance, as usual. Performance difference here is unlikely to be noticed in real world use.
5 GHz Uplink Throughput vs. Attenuation
Peak Wireless Throughput
For our peak wireless performance tests, the octoScope Pals are configured as 4x4 AC devices and left to negotiate their best connection, with 10 dB of attenuation applied on 2.4 GHz. The latter is necessary so the 2.4 GHz Pal isn't overloaded.
2.4 GHz Peak Wireless Throughput comparison - downlink
The AC68 does pretty well at 505 Mbps downlink and 339 Mbps uplink in these 2.4 GHz benchmarks. But since it's only three streams, it can't reach the same throughput as four-stream products, like NETGEAR's R7800 and ASUS' own RT/GTAC5300.
2.4 GHz Peak Wireless Throughput comparison - uplink
The 5 GHz peak benchmarks show the AC86U running closer to its four-stream brethren because it also is four stream.
5 GHz Peak Wireless Throughput comparison - downlink
911 Mbps downlink and 943 Mbps up mean you are essentially getting the equivalent of gigabit Ethernet performance. Keep in mind, this is with signal levels you're unlikely to encounter in real-world use.
5 GHz Peak Wireless Throughput comparison - uplink
I did not test MU-MIMO, because it generally doesn't provide significant performance advantages.
The ASUS RT-AC68U/NETGEAR R7000 "First Look" review has been in the top-five most read articles every week since it was published in October 2013 (!), an eternity in the Wi-Fi router biz. So the question is, is the AC86U (and NETGEAR R7000P) likely to now become perennial favorites?
Since it currently ranks #2 vs. the R7000P's #8 in the Router Ranker, the AC86 has more potential to become a favorite. The Most Popular Wireless Reviews - Trend page, which is updated weekly and shows the four-week trend of top 5 most read wireless router reviews, is where you can keep track. So far, it hasn't shown the R7000P to be particularly popular.
Instead, it's NETGEAR's R7800 four-stream Nighthawk X4S that seems to be getting steadier attention. But with the R7800 now ranking one step below the RT-AC86U's #2 spot and at around $30 higher, that might change soon.