Netgear EX6200 Hardware Revision Adds 2x2 MU-MIMO

[mediatrek]

A few days ago Netgear updated the EX6200 wireless extender product page and added the EX6200v2 support page.

The EX6200v2 hardware revision of the EX6200 wireless extender now offers MU-MIMO and is powered by a “quad core processor” versus the Broadcom “dual core processor” offered in the original hardware model.

While the FCC internal image documents (see PY316100335) are still confidential for the EX6200v2, going by the last two nuggets of information, one can guess that the EX6200v2 is based on Qualcomm chip solutions; specifically the IPQ40x8/IPQ40x9 ARM Cortex A7 fully integrated system-on-chip product portfolio. The only other product I am really aware of at this time with these chip solutions that is for purchase now is the Asus RT-AC58U.

For those who have been waiting for 2x2 MU-MIMO wireless bridge solution for entertainment centers, it sounds like we will finally start having options.

 

[thiggins]

To me, 2x2 MU-MIMO makes no sense. Rule of thumb for maximum simultaneous client support is Number of streams -1. So with two streams, # of simultaneous clients is 1.
Even if you remove the -1, maximum throughput gain is with two STAs.

MU-MIMO is turning out to be the 3D TV of Wi-Fi. The industry is moving on to multi-AP, at least for this year....

 

[L&LD]

Tim, you're right about MU-MIMO in a 2x2 implementation.

But to me, multi AP is the '3D TV' of WiFi (i.e. the 'mesh' implementations we've seen so far). Particularly when a single router is sufficient (true for most smaller homes/businesses/apts).

 

[sfx2000]

To me, 2x2 MU-MIMO makes no sense. Rule of thumb for maximum simultaneous client support is Number of streams -1. So with two streams, # of simultaneous clients is 1.
Even if you remove the -1, maximum throughput gain is with two STAs.

MU-MIMO is turning out to be the 3D TV of Wi-Fi. The industry is moving on to multi-AP, at least for this year....

2*2:2 MU can make sense with Repeaters, esp. if 2 or more are in use, as this will save airtime for the non-MU clients that might be attached to the repeaters.

Interesting.. obviously there is a technical case to do this, otherwise it wouldn't exist...

 

[thiggins]

nteresting.. obviously there is a technical case to do this, otherwise it wouldn't exist..

Not necessarily true. Consumer Wi-Fi has pushed forward with many things that end up with no real benefit for the consumer.

But to me, multi AP is the '3D TV' of WiFi (i.e. the 'mesh' implementations we've seen so far). Particularly when a single router is sufficient (true for most smaller homes/businesses/apts).

In the end, multi-AP is the way to go. It doesn't require special technology on the STA and more effectively addresses capacity issues.
Yes, they need to workout backhaul management. But it's a better path to pursue vs. higher link rates, MU-MIMO, etc. IMO

 

[sfx2000]

Not necessarily true. Consumer Wi-Fi has pushed forward with many things that end up with no real benefit for the consumer.

You've caught me on a thoughtful moment - so let me share some insight and informed opinion...

Devil's in the details - but a "repeater/extender" as a MU client raises interesting ideas...

In the past - a repeater/extender did just that - get a frame, pass it on.. and as such, since WiFi is TDD, to the end point on the far end, it wastes a lot of time...

Put two extenders/repeaters - it's even a bigger hit.

Taking a repeater (or a MESH station) into a MU context - I can push much more per frame, and concatenate far side Client data into a single MU frame, vs. taking multiple SU frames, thus saving airtime - which at the end of the day, determine both capacity and speed...

It does solve a few problems - since MESH does have more overhead, we have to find space inside the timeline to squeeze data in between, and here, MU might actually have a good application compared to having a Single AP/Multiple client STA scenario...

Who knows - consumer WiFi in many ways in on the long tail after Wave 1 - between BigHonkingRouters and MESHBoxes - there's plenty of room here, at least from a consumer WiFi perspective - the BigHonkingRouter market has pretty much played itself out, as an example - the AC1900 is probably the best band/buck return, and getting better every day in that class from a value perspective, and I'm guessing that the take up and sell thru on BigHonkingRouters is less - because once one has a AC1900 class - one tends to keep it, and new sales are either N-class new users, or replacing existing failed units.

So the vendors are pivoting towards other means, which is why Mesh (and Mesh type products) are getting more attention.

Mesh plus MU - nice way to expand things - there's still work to be done on the Router side, as many of the AC1900 class are weak there, and recent developments there are encouraging - e.g. Intel working to the Roqus folks, and I do see a resurgence coming in for dedicated Router (minus AP) - mainly due to regulatory restrictions on the Radio Side, and vendors/third parties wanting to continue to improve on the rest of the network.

Hmmm....

 

[sfx2000]

MU-MIMO is turning out to be the 3D TV of Wi-Fi. The industry is moving on to multi-AP, at least for this year...

I think MU in a single AP context is a checkbox feature - along with TurboQAM, VPN (which some do just to check boxes), and tri-band "clever connect".

Some of these provide value, and some don't - and some provide value that is under the radar...

consider SU-MIMO (prior to 11ac, just MIMO) - it was the key enabler for 11n - extra stuff there was beamforming and STBC/LDPC, along with more that 2 streams - LDPC and STBC were incremental improvements, and they did improve performance at a given range, and 3 stream augmented it - beamforming was a bit of a mess...

11ac - Wave 1 was really about Wider Channels and higher order modulations - and this was a huge benefit - so now we're left with more spatial streams and MU-MIMO (which requires beamforming) - so Wave 2 is really about looking in the parking lot for nickels in the building that 11ac built in a traditional sense of a Single AP and multiple clients - and testing on the main site pretty much confirms this.

In that context - I would agree - it's 3d HDTV, and the TV industry has the same problem - everyone bought an HDTV, and the product lifecycle says that sales will fall - so we see 3DTV, UHD, 4K, etc... marketing buzzwords that tempt folks to upgrade - but TV's last a long time...

Mesh plus MU - this starts to make more sense - again like I mentioned earlier above, the devil is indeed in the details, and I think the devil is likely borrowing on the future on that one - but MU does solve a problem that MESH has always had - MU saves airtime, so we can pack more data on each frame - not at a higher rate, but we can serve more clients in the same time allocation...

I get it... and it's not 3DTv...

 

[L&LD]

In the end, multi-AP is the way to go. It doesn't require special technology on the STA and more effectively addresses capacity issues.
Yes, they need to workout backhaul management. But it's a better path to pursue vs. higher link rates, MU-MIMO, etc. IMO

Multi AP (via wired Ethernet) is the way to go.

Mesh in everyone's home and business? That's creating an 'answer' to a problem that brings it's own issues instead to the table.

MU-MIMO combined with Mesh (as sfx2000 suggests) may be a way to solve some of the immediate issues within a single network. But I can't see it working well for wide scale use (ever).

 

[thiggins]

You've caught me on a thoughtful moment - so let me share some insight and informed opinion...

Thanks for the commentary. We'll see how it plays out.

 

[thiggins]

Multi AP (via wired Ethernet) is the way to go.

Mesh in everyone's home and business? That's creating an 'answer' to a problem that brings it's own issues instead to the table.

I agree APs with Ethernet backhaul is best. But if all homes had Ethernet where it needed to be, there wouldn't be all these alternatives (extenders, MoCA, HomePlug, mesh).

"Mesh" is not going to be for everyone. Just as BHR's aren't, or UBNT Edgerouters or Mikrotiks, etc. It's just another alternative.

A small % of Wi-Fi consumers have the know how to set up a single router, let alone a multi-AP system. So the push to make multi-AP systems self-driving, which is also a part of "mesh" products, will have benefits regardless of the backhaul mechanism.

 

[sfx2000]

Thanks for the commentary. We'll see how it plays out.

Sounds good - more I think about it - MU is about capacity and efficiency - not outright speed - and with Mesh, it's a great chance for MU, as the big hit are those two items - and Mesh node typically are fixed - e.g. they don't move around inside the LAN, so sounding and ranging becomes much easier to work with.