RT-AC88U

[kloxxe]

The limitation may not be the router but the environment.

Indeed, but in the same environment I would expect the AC3100 to perform better than the EA6900 (which is a two years older design and costs half of the money).

Anyway...

My 100mbps internet is finally on and I see no difference in throughput between the EA6900 and the RT-AC3100. I get the same 70-80mbps @ 2.4Ghz and 97-100mbps @ 5Ghz (I guess at 5Ghz the limiting factor is the internet connection, not the wireless interface)

Based on above, and because the EA6900 costs less than half than the RT-AC3100, I will return the RT-AC3100 to Bestbuy tomorrow.

The RT-AC3100 may be a better router, but in my application (sharing 100mbps internet connection in a 3 bedroom condo for 3 wired and up to 10 wireless clients) it has no advantage over the much cheaper EA9600. Granted it has some advanced traffic measurement, QoS, etc.. functions, but again the EA9600 does what I really need, so I don't feel paying the 220 CAD extra.

BTW, The ASUS RT-AC3100 is on boxing day sale until December 31st in BestBuy Canada at 319 CAD (231 USD). The Linksys EA6900 is on sale there, too at 149 CAD (108 USD).

 

[sm00thpapa]

So I made some comparison test about signal strength between the EA6900 and the AC3100, by swapping the routers at the same location, same environment and monitoring the signal strength with Wifi Analyser at the worst covered spot in my condo.

There is no difference in radio signal strenght at all. Both unit provided -50dBm @ 2.4 Ghz, and -70dBm @ 5Ghz (both with +/- 3 dBm margin).

This was surprising, considering that the AC3100 has 4 antennas, each twice the size of the three EA6900 antennas...

Proving the point that size doesn't really matter.... ;-)

My 100mbps connection upgrade will be activated tomorrow, then I will do some throughput test too (only interested in sharing the internet connection, so if the EA9600 provides 100mbps throughput at that worst spot, the game is over and AC3100 goes back to BestBuy)

Just make sure you're testing with a 3x3 device. Just got my daughter an ASUS laptop and it came with a cheap 1x1 N card. So it only pulls 50Mbps of my 200Mbps connection.

Sent from my LG-D850 using Tapatalk

 

[kloxxe]

I won't get more than 100mbps (limited by the 100/10 internet connection), so for my case it does not matter if the wireless interface could go beyond that. Also my clients are given, some smartphones (Samsungs mainly), a Chromecast, 3 wired PC and some laptops (the latest ones are an Asus ROG and a HP Folio (both brand new, Core i7 6th gen units). All of these have Wireless AC receivers, bot none of them are capable of receiving beyond ~800mpbs even at ideal conditions...

I'm sure in a testlab the AC3100 would perform much better then the EA9600. It has to. In my specific case these advantages are not there, mainly because both routers outperforms the other limiting factors (100mbps connection, wireless clients).

 

[L&LD]

Before returning, it may be interesting to test the two routers by manually adjusting and testing for the best channel for your environment.

For the 2.4GHz band, test channels 1, 6 or 11 (and 13 if supported in your country) at 20MHz width.

For the 5GHz band, test each channel at 40MHz width (as none overlap at that width).

Specifically, test for the responsiveness of web surfing on each channel (at one or more locations) and the throughput you can achieve from a wired computer to your test laptop in MB/s (the higher and the least variability, the better) with a single large (1GB or larger) file. Of course, you can also test using ookla speedtest too, but do this in addition to the above.

With the Asus, you may also get results like the following with future firmware from third parties like hggomes, for example.

http://www.snbforums.com/threads/asuswrt-merlin-378-55-3_hgg-final-mod.26524/page-2#post-199549


While you can save $100 today by returning the RT-AC3100, I think that is a small cost over the time you'll have use of the router when compared to the potential it may offer you in the near future.

Of course, don't buy or decide for what 'might' happen. Do the testing above and see what an extra $100 gets you now.

 

[Mordred]

Looks like the SOC in the AC88U/R8500 is crap:

http://www.dd-wrt.com/phpBB2/viewtopic.php?t=287177

The new broadcom draws too much power compared to the qualcom , no wonder they all come with huge heatsinks.

The EA8500 has a tiny heatsink:

http://www.smallnetbuilder.com/wire...00-mu-mimo-smart-wi-fi-router-reviewed-part-1

obviously because the cpu architecture in the broadcom is outdated.

 

[thiggins]

Kloxxe: I think you've done your due diligence. Go ahead and return the router. In a year (even less), if you want to upgrade, there will be plenty of other choices, including these and other AC3100 class routers that will be much cheaper.

If does not pay to be an early adopter of new wireless routers. The smarter buy is to stay a generation (or two) behind.

 

[punkinduster]

I'm thinking of sending my AC-88 back. I found what looks like a firmware bug, acts like a firmware bug, smells like a firmware bug, BUT no one else acknowledges seeing it except for me. Even RMerlin hasn't responded as to if it might be a firmware or hardware problem. I don't mind waiting for a firmware fix if I knew it wasn't just MY hardware problem. I guess it is a low priority since it is just internal graphs for real-time global traffic data and the actual data flow seems to be working just fine. I'm guessing it has to do with the extra 4 ports on the AC-88. Thanks for listening. **rant off**

 

[kloxxe]

Hmmm... There are a couple of things I just noticed, in favor of the AC3100:

- Whenever I activate Media Prioritization on the EA6900 (I want to give priority to my Chromecast over my kids gaming and downloading), the speed goes down by 10-20 mbps, even the Chromecast doesn't do anything. It seems the EA6900 reserves part of the bandwidth whether there is a need or not.

The RT-AC3100 Adaptive QoS seems to work much better, I have assigned high priority to the Chromecast and set the automatic settings to Media Streaming instead of gaming. The speed (measured by speedtest.net on one of the wired PC) was not affected.

- Another thing I actually found quiet useful. When I hover the mouse above the connection signal in Network Map/Client Status, it gives me the radio signal strength and the send/receive throughput. No need to play around with a phone and Wifi Analyser anymore...



- I kind of like the detailed traffic analysis, too (seeing where my bandwidth goes ), the EA6900 lacks any sort of option for that...

Hmmm... I think I keep testing it today before the I make my final decision.

 

[faur]

Hmmm... There are a couple of things I just noticed, in favor of the AC3100:

- Whenever I activate Media Prioritization on the EA6900 (I want to give priority to my Chromecast over my kids gaming and downloading), the speed goes down by 10-20 mbps, even the Chromecast doesn't do anything. It seems the EA6900 reserves part of the bandwidth whether there is a need or not.

The RT-AC3100 Adaptive QoS seems to work much better, I have assigned high priority to the Chromecast and set the automatic settings to Media Streaming instead of gaming. The speed (measured by speedtest.net on one of the wired PC) was not affected.

- Another thing I actually found quiet useful. When I hover the mouse above the connection signal in Network Map/Client Status, it gives me the radio signal strength and the send/receive throughput. No need to play around with a phone and Wifi Analyser anymore...



- I kind of like the detailed traffic analysis, too (seeing where my bandwidth goes ), the EA6900 lacks any sort of option for that...

Hmmm... I think I keep testing it today before the I make my final decision.

...for me, another + is the presence of VPN... as far as i know, there is no VPN on Belkin EAxxxx...

Sent from my SM-G900F using Tapatalk

 

[punkinduster]

I have the proper graphs - note that I have the RT-AC3100, so if you suspect the issue is with the 4 additional LAN ports, my result does not count. Can you plug your cable in Port1-4 and check if you get the Traffic report correctly from there? That would confirm the problem with the 4 additional ports...

I moved the wired ports around and that seems to be working OK. What is strange is I can run an Internet speed test with a 2.4 wireless connection and the graph will appear on the Wired tab and show as double the transfer rate of the ISP connection on the WAN tab graphic. This is with Merlin's builds not 380-858.

 

[kloxxe]

This is with Merlin's builds not 380-858.

I am no expert here, but I'd try to load back the latest factory firmware (Version3.0.0.4.380.858) and check if that is ok...

 

[punkinduster]

I am no expert here, but I'd try to load back the latest factory firmware (Version3.0.0.4.380.858) and check if that is ok...

That is where I am now. The graphs are better. Only a couple of issues. WAN graphs show appropriate data. The 2.4 GHz data doesn't appear on the Wired tab. The 5GHz receiving graph data does always shows zero but the transmit graph is OK. This is why I assume it is a coding issue. 380-858 works better than Merlin's firmware as far as these graphs are concerned but I really like what Merlin has added to the mix. My only concern at this time is no one else has seen these issues but me.

 

[faur]

That is where I am now. The graphs are better. Only a couple of issues. WAN graphs show appropriate data. The 2.4 GHz data doesn't appear on the Wired tab. The 5GHz receiving graph data does always shows zero but the transmit graph is OK. This is why I assume it is a coding issue. 380-858 works better than Merlin's firmware as far as these graphs are concerned but I really like what Merlin has added to the mix. My only concern at this time is no one else has seen these issues but me.

...and you think that EAxxxx will give you better graphs...or better infos/options?! Anyway...it's your choise...

Sent from my SM-G900F using Tapatalk

 

[punkinduster]

...and you think that EAxxxx will give you better graphs...or better infos/options?! Anyway...it's your choise...

Sorry? I think you quoted the wrong guy. I'm just trying to figure out if my RT-AC88 is broke or not.

 

[faur]

Sorry? I think you quoted the wrong guy. I'm just trying to figure out if my RT-88 is broke or not.

....you're right...sorry...


Sent from my iPad using Tapatalk
Mikrotik(for routing) + AC88U(for 5Ghz) + AC68U(for 2.4Ghz) + Cisco switch(for NAS link aggregation)

 

[kloxxe]

- I kind of like the detailed traffic analysis, too (seeing where my bandwidth goes ), the EA6900 lacks any sort of option for that...

 

[kvic]

Looks like the SOC in the AC88U/R8500 is crap:

http://www.dd-wrt.com/phpBB2/viewtopic.php?t=287177

The new broadcom draws too much power compared to the qualcom , no wonder they all come with huge heatsinks.

The EA8500 has a tiny heatsink:

http://www.smallnetbuilder.com/wire...00-mu-mimo-smart-wi-fi-router-reviewed-part-1

obviously because the cpu architecture in the broadcom is outdated.

The CPU in the newer ASUS routers performs as expected without any surprise. It's very similar to the CPU in older models (AC56U/AC68U) but clocked at 1.4GHz. An overclocked AC56U (at 1.4GHz) can achieve similar numbers:

type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 bytes
aes-256 cbc      34005.28k    36591.54k    38551.66k    38169.77k    38291.35k

But not to forget that the BCM4366 WiFi chip in AC88U comes with its own co-processor dedicated for WiFi related processing (some source says at 800MHz). This offloads lots of work from the main 1.4GHz CPU.

Some folks reported earlier in this thread. During 802.1ac throughput test at ~500Mbit/s, both cores on AC88U are barely utilised, between 1% to 5%. Put it into perspective, at the same throughput AC56U (o/c 1.4GHz or if it matters AC68U), CPU 1 is >50% utilisation. The benefit of newer wifi chip is clear.

On the contrary, I'm disappointed at openssl benchmark of R7500/EA8500 at 1,4Ghz. They come with Qualcomm Atheros SoC. According to the benchmark data from your link, performance is not categorically better than Broadcom. Not to forget that there is h/w AES accelerator in there.

http://images.anandtech.com/doci/7526/IPQ8064.png

Nevertheless you're right that Qualcomm SoC's features are very appealing and competitive...as long as firmware can make good and full use of.

 

[red_pope]

I suspect it might help mostly for people connecting to oversea game servers, as these might provide with better routes than what their ISP uses. But someone in the US connecting to a US-based game server, I doubt it.

I will eventually give it a try with that one online game I play. My average ping time is between 90 and 150 ms, as the game server is quite some distance away. Will be easy to see what kind of latency difference it makes. I just need to wait for the feature to be fully implemented however (it's not complete yet in 9177).

RMerlin

Redirecting traffic is old school stuff. Yet you still have to go through your ISP first. Is all about DNS servers, unless you are going through a VPN. I honestly love the modern looks and features but for know Ill stick to my RT-AC87U. Maybe Ill be tantalizing, if the price is right,

 

[kloxxe]

Anyone knows how to edit the client list for removing already disconnected clients? I moved one of my PC from wired to wireless and at "parental control/time scheduling/client list" I now have two entries with the same name (bit different MAC addresses???). Since it only shows the name and the MAC address, I would really like to remove the old entry...

 

[CyberAthlete]

I would not try to "future proof". You're overpaying for things that, in the end, will be replaced by other technology coming in less than a year.

If you want to improve your wireless network, buy another AC1200 or AC1900 router, convert it to an AP and connect it via Ethernet, Powerline or MoCA to the first one and place it at the other end of the area you are trying to cover.

I agree with you and I posted this in another forum elsewhere (PC's and TVs) that future proofing is not a smart investment. By the time that future is here, better and cheaper hardware will have replaced what you buy today. So might as well save that money and buy the things when you actually need them. Then you're paying for a better product (for the same or even less). And these products will be a lot lot cheaper too.