Orbi vs Nest vs Eero Pro - Needing assistance on purchase

[Tech Focus]

Yes. That's correct, and the stability just make whole experience far better. Technically speaking single client device speed, 2x2 and 4x4 router shouldn't make difference as long as client is 2x2 but with multiuser MIMO, 4x4 theoretically should allow one to have TWO 2x2 client at max speed simultaneously. But again, how often we use max connection speed. Plus if you have two maxed 2x2 clients, for internet that would exceed 1 Gbps. So internet would become bottleneck.

As far as roaming (switching) as i recall Eero was smooth and so is UniFi, but neither is perfect and I don't think there is any system that is perfect meaning it won't be always switching to the fastest/nearest nodes immediately. If I run continuous WiFi speed test and walk around, you can see node switching i.e. a few seconds drop of WiFi but doesn't really affect what we do day to day i.e. it's not like you lose Skype call while moving around home. UniFi supports 802.11r,k and supposedly v whatever they mean..(supposedly fast roaming helper protocols). With quick google, I'm not sure if Eero supports these standard. But I also don't know how much these standard make roaming smoother or faster.

As far as coverage, apparently enterprise APs tend to have lower coverage per device than consumer product due to the fact enterprise tends to put more AP. Having said that for our 3800 sqft (ground & basement level), Eero Pro x 3 covered essentially all the living space with some area speed going well below 50 Mbps but worked except one end of home but that area I didn't care WiFi coverage because it's a basement laundry room/storage room surrounded by brick. With UniFi, one HD & one nanoHD covered entire home end to end including the area that wasn't covered before BUT UniFi I have it with wired backhaul and Eero was Wireless. In fact, I just checked. Currently, only 3 devices are connected to nano-HD (located one end of home) and everything else on HD (middle of home) and none on IW-HD (I really didn't need this...).

Basically between Eero Pro and UniFi, the reason why I would personally choose UniFi are followings:
- Scalability: When UniFi release WiFi 6 or WiFi 6E AP, I most likely want to buy one. I can still use any of AP I currently have. Just provide more area of 5 GHz coverage, or may be I will just sell one unit, and slowly replace one at a time who knows.
- Customization: I tries with Asus when I got it before but somehow it wasn't as fun (may be interface) but I like UniFi interface even some feature may be called "gimmick" by some. For instance, it has heat map to let me put virtual device as well as one I own over the my floor plan and decide where is optimal placement, if I need new AP. WiFi AI to let it choose channel every night, multiple SSID on same APs (I have smart connect SSID & 5G only SSID, I always wanted this) and who knows what else I discover tomorrow.
- Security: I hate subscription even if it turned out cheaper than paying life-time license. UniFi have no subscription intrusion detection/protection. This actually completely stopped warning on my NAS.
- Hyperactive community: UniFi community forum is super active. If I put comment, I get reply within 10-15 minutes.
- Speed: I have 3x3 MIMO capable laptop so Eero won't do it.

Again, I may be at the stage of just started relationship (not seeing anything bad) but I don't think I've ever been this much excited about network system before. For me UniFi just hit every sweet spots.

 

[lardo5150]

Kind of off topic, but if you had either the Eero Pro with two wired up and one not (which will be my setup), or just two Unifi's, would you name each AP the same SSID? I always got confused on this.

 

[Tech Focus]

Eero I believe you may not give you even a choice but single SSID as its Mesh system.UniFi, you can name anyway but if you want to have just smooth single WiFi, which I believe you do i.e. walk from one end of home to another without physically selecting new SSID, you just have one SSID. People talk about separating 2.4 GHz and 5 GHz bands. That's separate.

 

[Trip]

A few clarifications/additions:

with multiuser MIMO, 4x4 theoretically should allow one to have TWO 2x2 client at max speed simultaneously. [...] if you have two maxed 2x2 clients, for internet that would exceed 1 Gbps. So internet would become bottleneck.

Theoretically, maybe yes, but in the real world, typically no, for both. MU-MIMO is not actually field-usable by most (if any?) client hardware, still to this day, despite many labels indicating otherwise (@thiggins can corroborate if that's still the case). Also, real-world max for a 2x2 client on 5ghz 80Mhz channel width (160 isn't all that practical or supported) is 866Mb/s, but it's more like 2/3 of that at the most, due to TCP/IP overhead, so ability to max a 1Gb internet link (at least download) is a potential, but likely not a certainty.

I don't think there is any system that is perfect meaning it won't be always switching to the fastest/nearest nodes immediately.

Presuming the clients and their chipsets/drivers properly support advanced roaming, there certainly are systems that can create truly seamless roams, from both the standpoint of the user and the packet order. And it's not necessarily cost-dependent. I haven't confirmed, but TP-Link Omada may actually have a better 802.11k/r/v implementation than UniFi (I know it's been a bone of contention for years now on the UI forums...). There's also Cisco WAP, for which @coxhaus has had really good results on when hard-wiring several WAP581's. Beyond that of course you've got full enterprise products like Ruckus, Aruba, Cisco Aironet, etc., which offer the highest levels of seamless-ness and roam crafting, but of course tend to carry hefty price tags.

As far as coverage, apparently enterprise APs tend to have lower coverage per device than consumer product due to the fact enterprise tends to put more AP.

As a general trend when looking at amplification, often yes, but engineering for distributed wifi also involves a lot of spatial density planning as well, which relates to the actual RF and antenna design capabilities, as much as it does to raw amplification levels.

For example, I recently extended a customer's wifi to his garage/shop (a building of maybe 30 feet by 60 feet) via Cat6a, a Cisco Catalyst switch and a Ruckus AP. I surveyed with both an R510 and an R710, expecting the 710 to show superior survey results... when just the opposite happened. The 510 was better in nearly every regard, even though it's lower-spatial stream, lower power and lower client capacity; the actually reason, I think, is proximity tuning, the R510 being tuned closer, the 710 further away (example: 100-foot lecture hall ceilings). So, as the old adage goes, site surveys often can't be beat, if you can afford for them.

Great additional feedback, though. Nice to have more people as data points on UniFi, Eero and the like.

 

[L&LD]

Just a data point. I'm running 160MHz bandwidth with no issues and much higher throughput on the RT-AX88U and a mix of AX and AC clients.

 

[Trip]

Just a data point. I'm running 160MHz bandwidth with no issues and much higher throughput on the RT-AX88U and a mix of AX and AC clients.

That's awesome, and great that 160Mhz is usable in your airspace. Do you have much neighboring wifi to contend with at all, or ever?

 

[L&LD]

Lots of WiFi, but far (enough) off.

 

[Tech Focus]

@Trip

Thanks for clarifications.

MU-MIMO is interesting as I just going off by what I read and conceptually make sense, but I must admit, I don't know if I ever tried to test this nor feel like experienced the benefit myself. For AiMesh I had to turn it off for better stability.

Overhead is certainly something I learned during my forth (most recent) round of network system search, and was really surprised 30-50% speed loss is inevitable. Also, MIMO and 160 MHz concept were so crucial for understanding WHY 4800 Mbps WiFi 6 is not really what we see as a user. Because I used to think full vs. half duplex, 4800/2 = 2400 Mbps that's still twice more than 1 Gbps wired connection.

I believe iOS has been supporting k/r/v for awhile but if I slowly walk while WiFi analyzer app turned on to measure speed and when I see a slowdown in speed, and stop right at the location, I can actually see disconnect/stop analyzer while if keep walking the analyzer will continue to run and switch to another node. So why do I do intentionally stop? Well... no real reason. But yes, practically, I see my device roaming is seamless on UniFi for my home (but I can break it).

@L&LD

I've never give a try to 160 MHz as I read it's usually not functional, but may be I should. That's double the speed.

 

[L&LD]

@Tech Focus I don't know if 160MHz bandwidth is usable for you and your router + client devices, but that was part of the point of my first post. Everything must be tested before a conclusion can be reached. And even then, only reached for that one specific network.

 

[avtella]

MU-MIMO actually works decently on the likes of the R7800 and similar Qualcomm QCA9984 based routers, I have 5 MU capable (3x AX200, 1x QCA6174A aka Killer 1435, 1x8265) devices and I saw a nice gain of course not night and day but still decent in simultaneous throughput. Infact if I recall SNBs own tests showed the R7800 was pretty decent with MU-MIMO. MU is most effective when devices are a decent distance apart.

However Broadcom AC routers and client devices were pretty terrible actually causing performance loss at times with MU enabled, maybe the new Broadcom chips are better. Most of the bad MU rep seems to stem due to Broadcom. I mean yeah it’s not a make or break feature but can be nice to have.