What's new

GT-AX6000 vs AXE7800

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

MrErik

New Around Here
Hi everyone, wanted to get your advice on getting the GT-AX6000 ($300 USD) vs the AXE7800 ($330 USD).

Currently using a Synology RT2600AC but want to upgrade for better speed and range. Using 2x1000 internet connection and no 6E devices. The price difference between the two is not much so originally I was thinking of getting the AXE7800 for futureproofing (6E). However, I'm wondering whether it makes sense to go with the AX6000 for the 4x4 2.4GHz, slightly faster CPU and more RAM. What does everyone here think?
 
So there is zero point wasting @$300 on anything. If your Synology router is working stay with that for now. You can buy extra turkey and beer for Christmas :p

WiFi 7 is coming so buying 6E routers is not future proofing.

6E is currently pointless on user client end.

Only real world benefit is wireless backhaul (valid in dense areas) and early adoption intel clients (no/low interference). Most 6E hardware is low power via FCC rating.

Sorry, I don't consider phones to "need" 1gbps+ download speed at home..

Industry usually favors cheaper hardware, hence why 5G WIFI 6 hasn't really broken through. A lot of IoT stuff still falls into the AC/N hardware categories.

I believe it's viewed as diminishing returns at least until pricing (for AX) is more in line with older HW. 2x2 AC> 1x1 AX obviously.

WIFI5 to WIFI6 isn't that much different when you compare top end hardware designs, granted WIFI6 hardware IS faster in terms of overall internal radio speed and client capacity. Real world performance depends on the HW itself.

Is WIFI6 better? Yes.. I can push higher speeds at similar distances with interference (25-30FT), but its mostly modulation/SNR improvement when comparing similar bonding via same 2x2 client at fixed distance. (similar router design IE: AC86>AX86)

All top end 4x4 BCM (ACW2/AX) designs will end up with a 600-800mbps max speed threshold at 80mhz bonding via 2x2 client close range.

160mhz bonding lets certain 160mhz capable clients (Intel AX2xx/4xx for example) break that limitation.. That includes AC W2 with 80+80 across the two 160mhz blocks.

The only way to get over 600-800mbps on 80mhz bonding is 3x3 or 4x4 MIMO between main router.. AX does better than AC here, granted there are no 3x3 or 4x4 clients on market... Only valid between 2 APs.

BCM 4x4 ACW2 > 4x4 ACW2 80mhz only peaks at 800mbps. 2x2 client on same AC HW is around 600mbps.
 
Last edited:
Not much for a link between two routers in AiMesh:

AC86U to AC86U -> 2166Mbps*
AX86U to AX86S -> 2400Mbps

* - non-standard 1024-QAM support

Correct, but throughput is a lot higher

4x4 AC86 > AC86 80mhz only peaks 800mbps. Even close range.

4x4 AX86 > AX86 80mhz will push over 1GBPS easily.
 
Last edited:
but throughput is a lot higher

No, it's not. AC86U with standard 256-QAM and 1733Mbps link speed in Media Bridge mode does 940Mbps limited by the Gigabit ports. Absolutely the same as any AX router with Gigabit ports. Tested with AX86S. Even Wi-Fi added latency is the same. You like theoretical statements, I like testing.
 
No, it's not. AC86U with standard 256-QAM and 1733Mbps link speed in Media Bridge mode does 940Mbps limited by the Gigabit ports. Absolutely the same as any AX router with Gigabit ports. Tested with AX86S. Even Wi-Fi added latency is the same. You like theoretical statements, I like testing.

Im basing that on a 4x4 PCE-88 (4366 non "E") client with the same 1733MBPS/2166 nitro/turboQAM link speed. It would never budge over low 800mbps, even close range.

Modem throughput limitation? PUMA7 would always give me worse wireless throughput (with ACW2) even though they would peg the same hardware limitation wired. I cant test this now as I downgraded my ISP plan a tier below (500mbps rated, 580mbps over provisioned), but I do have a newer BCM3390 based modem provided via ISP.

It's possible that AC86U >AC86U can do full 940-950, in which case maybe I'm wrong.


For AX86, I was referring to U variant. Wireless backhaul/media bridge with same hardware + 2.5G port could push over 1GBPS easily.. granted the media bridge would/could force 160mhz (US variant) unless a UNII-3 channel was set manually.


You like theoretical statements, I like testing.

Yet you're consistently recommending low end 2x2 HW to people expecting it works the same way for everyone that it does for you. Interesting..

I'm not saying 4x4 does better for everyone either, but I would take the higher end MIMO hardware 9/10 times, especially in dense environments. Every 2x2 AX solution I've tried has been absolute garbage in my 1700 SQFT home. Thats my bias.

I also test most stuff from subjective experience/environment and will make it clear that WIFI is RNG even down to modem selection. It's obvious you don't like being challenged in this area. :(
 
Last edited:
6E is currently pointless on user client end.

Only real world benefit is wireless backhaul (valid in dense areas) and early adoption intel clients (no/low interference). Most 6E hardware is low power via FCC rating.

Sorry, I don't consider phones to "need" 1gbps+ download speed at home..

Industry usually favors cheaper hardware, hence why 5G WIFI 6 hasn't really broken through. A lot of IoT stuff still falls into the AC/N hardware categories.

I believe it's viewed as diminishing returns at least until pricing (for AX) is more in line with older HW. 2x2 AC> 1x1 AX obviously.

WIFI5 to WIFI6 isn't that much different when you compare top end hardware designs, granted WIFI6 hardware IS faster in terms of overall internal radio speed and client capacity. Real world performance depends on the HW itself.

Is WIFI6 better? Yes.. I can push higher speeds at similar distances with interference (25-30FT), but its mostly modulation/SNR improvement when comparing similar bonding via same 2x2 client at fixed distance. (similar router design IE: AC86>AX86)

All top end 4x4 BCM (ACW2/AX) designs will end up with a 600-800mbps max speed threshold at 80mhz bonding via 2x2 client close range.

160mhz bonding lets certain 160mhz capable clients (Intel AX2xx/4xx for example) break that limitation.. That includes AC W2 with 80+80 across the two 160mhz blocks.

The only way to get over 600-800mbps on 80mhz bonding is 3x3 or 4x4 MIMO between main router.. AX does better than AC here, granted there are no 3x3 or 4x4 clients on market... Only valid between 2 APs.

BCM 4x4 ACW2 > 4x4 ACW2 80mhz only peaks at 800mbps. 2x2 client on same AC HW is around 600mbps.


Was there a point to your rant?

I merely pointed out to the OP that it was pointless buying the expensive routers when he states he doesn't use any 6E kit and that as wifi7 is coming he wouldn't be "future proofing".

Your lecture was really not required.
 
Currently using a Synology RT2600AC but want to upgrade for better speed and range.

Well - RT6600ac is a good choice - feature set is similar to your RT2600ac, and you can use the RT2600ac as a mesh node once you have the RT6600ac in place.


The Syno's are decent units...
 
I don't know about RT6600ac, but I wouldn't recommend buying AX86U today. It has firmware issues lately, already replaced by newer hardware AX86U Pro and perhaps going to be discontinued soon. If you want an Asus router with expected longer support go for GT-AX6000 or wait for AX86U Pro.
 
You don't recommend the superior hardware ever. Yet you do recommend the crippled 's' version for a savings of a handful of dollars.

No firmware issues in many deployed networks. Not replaced by a longshot by any (name only) 'pro' model. Your conclusions are not based on facts, nor any direct real-world experience with the equipment.
 
I personally think the "crippled" AX86S is fine with current US price gap of $70 (or $100 with best buy trade in), but I would never recommend it over AX86U if pricing was closer to the MSRP.

I can't confirm the exact spec, but the older BCM43684 4x4 SoC (AX86U/S 5G radio) has a fairly beefy internal A7 core (1.5Ghz?) that doesn't need to offload to main CPU as much as older/alt designs, granted that also means that the AX86U and PRO model could handle "more" for obvious reasons.

As for the 2.4G 3x3 BCM6710 chip, it does impact the main CPU (to a noticeable degree) when running a basic speed test, but I can see that being logical as the chip is fairly small relative to the 4x4 Radio SoCs. I just have no clue what the internal spec is or if it shares an A7 core like the BCM43684 (and BCM6715.. 1.7Ghz?).

I don't think dual core BCM4906 via AX86S will ever get heavily saturated on 5G network, especially in a 1-2 person home, though the RAM is a bigger culprit all around, especially with newer ASUSWRT 388 firmware code. AX86S on current FW consumes 375-400mb out of box with nothing enabled... I would imagine the bottleneck to start here as time goes on.

I really wish there were white papers or marketing material like Broadcom used to let out.. IE: the 4x4 BCM4366/E is a confirmed 800mhz A7 SoC design as referenced here via "penta core" design:


At least the dual core A53 4906 in AX86S is better than the lower end full SoC "tri/quad core" designs.. Which are more or less confirmed 1.5ghz and 1.7ghz A7 for Gen 2. CPU pool is shared between radio... Marketing nonsense.

I assume BCM43684 is a 1.5ghz A7 due to 1 vendor sort of leaking that spec, but it also makes sense relative to the "real" SoCs found in entry routers.

I'm very curious about the 3x3 6710 radio for that reason.. Would be weird if its weaker clock spec or if it fully offloads to main radio with minimal use.. Would rather have the full 4x4 + 4x4 design for that reason alone.

Sorry for rant.
 
Last edited:
I didn't see that as a rant. Thanks for your perspective.

My perspective isn't limited to the single moment we put our money down on the counter to purchase one or the other.

I'm basing this on the 5+ years of future use (ours or the next owners after we give/sell it away), and the 2 years past ownership that I've had zero issues with the specific model (RT-AX86U).

Saving $70 sounds good in theory. Saving $14 or less (if we keep routers longer) over the next few years is 'overkill' in the wrong direction, to me.

I'm from the era when you buy quality once, to save money (over a given period). Not, buy based on an arbitrarily lower dollar sum (repeatedly), yet effectively squander your funds while getting worse network performance too.

Just as things should be made as simple as possible, but not simpler...

Purchasing enough hardware, at a fair current price, is the better path forward. Always.

If our budget doesn't allow this to happen today, we keep saving until it does.
 
I agree, but it's a matter of perspective and end goals. WIFI used to move a bit faster than it does now, especially on client end. Case and point 6/6E marketing vs actual clients.

We can look at the AXE7800 the OP mentioned. Yes its a "Quad Core", but they're 1.7ghz A7 shared cores with the same 512mb ram pool... And thats a $329 USD new release. Same case for the mesh GT6 @ $600.

If anything, the dual core A53 4906 based routers have more Omph relative to full SoC variants with same ram pool if were factoring the main radio, granted, the firmware on AX86S is shared with AX86U and might end up with its own unique issue as time goes on.

Edit: That's also not saying that the lower end A7 SoC aren't enough for some people.. The WIFI portion of the router is going to be as strong as the components (amps/FEM, EE design) and radio MIMO config used.

AXE7800 with its 4x4 BCM6715 5G radio might not need to dip into the main shared 6756 SoC.. but we're aware that A53 based options are innately better as a backend CPU.

I guess that's my argument... at least when it comes to how Broadcom hardware seems to present itself and operate.
 
Last edited:

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top