Should the SNB standard test client support 256 QAM in 2.4 GHz?

[sfx2000]

And FWIW - that's also why one can mix MCS values in MU - QAM16/64/256 are all orthogonal - so can overlay a QAM16 on top of a QAM256 and it works

I know, because myself and a team of others did research on this back in 2009 for Wimax/802.16

 

[RMerlin]

They might have skipped it. The webui isn't very detailled...

 

[tim]

I would be very interested to see if QAM256 (on 2.4Ghz) actually works.

Could be a failure, could be enlightening. Still would like to see the results.

 

[sfx2000]

I would be very interested to see if QAM256 (on 2.4Ghz) actually works.

Could be a failure, could be enlightening. Still would like to see the results.

It does - but it's very dependent on environmental situation - most folks likely won't see any difference, as when in Turbo, there's a lot of retransmissions and bad frames due to noise...

 

[sfx2000]

And with most clients - they don't support it anyways - some might, but in VHT20, best thruput is AC289 vs. N216 with a three stream client, and that N216 will be much more robust and reliable.

In 2.4GHz, best to stay with HT20 (B/G/N mode), and leave VHT20 to the spec sheets...

 

[sfx2000]

Also interesting to note that even Broadcom says that TurboQAM might be up to 10 percent faster than 11n in 2.4GHz... doesn't sound like much benefit vs. the interop risks..

• Broadcom's TurboQAM® technology implements the highest data-rate 256-QAM mode in 2.4 GHz and enables devices equipped with the BCM4335 to deliver 10 percent faster throughputs than 802.11n speeds when communicating with other 5G WiFi devices

http://www.broadcom.com/products/Wireless-LAN/802.11-Wireless-LAN-Solutions/BCM4335

My thoughts are that WiFI Alliance could put a halt to these "proprietary" extensions by threatening to remove their certification - wouldn't be the first time...

Same goes with the QAM1024 implementations in 11ac for 5GHz... (and 2.4Ghz)...

Those who were around back in the 11b/11g days probably recall all of the interoperability issues back then, and how it really impacted the user's experience... Super-G, Afterburner, Turbo108, Nitro, etc... really would hate to go back to those days, and it's my thoughts and opinion that SNB can take the high road and not evaluate it...

 

[sfx2000]

Here's an eye chart of all the different things that happened in the latter days - notice most of them, to make best use, ended up getting into a vendor lock-in - otherwise, 11g, and sometimes much worse due to competing tech...

(also a reason why it took years to get 11n done - so many things went into that crockpot of standards... MIMO made it out, frame aggregation made it, channel bonding made it, but it took hella effort on the part of TGn to get consensus...)

Higher Data Rates;

Longer Range;

 

[thiggins]

Thanks for the charts and the history lesson, sfx.

Airgo was truly impressive when I tested it back in 2004.
http://www.smallnetbuilder.com/wireless/wireless-features/24291-doextended

 

[sfx2000]

Thanks for the charts and the history lesson, sfx.

Airgo was truly impressive when I tested it back in 2004.
http://www.smallnetbuilder.com/wireless/wireless-features/24291-doextended

Airgo was quite good, and that DNA is still around... I was alway fond of the Video54 solutions - which is now Ruckus...

 

[Razor512]

When doing a wireless bridge, it has been hard to actually achieve 256 QAM since it seems to struggle with interference, but it seems worthwhile to test, especially to see how well some routers do with it.

 

[sfx2000]

Hi Tim - just wanted to send over a quick note - looks like Intel isn't supporting Turbo/NitroQAM in their drivers on their AC cards - I've got a couple of them - a 7260 and a 7265 (AC models confirmed) - looking at Wireless captures for these cards on the Probe Requests, they're vanilla 802.11n... 2*2:2 - 5GHz, again, pretty vanilla, SU beamformee support, no MU, no 80+80/160 channel support.

No "special sauce", e.g. no Vendor Specific Attributes - intel remains one of the cleanest implementations out there for 11n/11ac - at least with the current drivers as of 4/29/16

 

[thiggins]

Thanks, SFX. I don't think the newer 82xx series supports it either.

 

[sfx2000]

Thanks, SFX. I don't think the newer 82xx series supports it either.

Yeah, and they're pretty much sticking with 2*2:2 on their upper tier cards, even with their 2016 roadmap - wonder if this is a power concern with NGFF?

 

[sfx2000]

Thanks, SFX. I don't think the newer 82xx series supports it either.

Not sure if anyone else saw this, but while the AC8265 doesn't support QAM256 in 2.4, it does appear to support MU-MIMO...

http://www.intel.com/content/dam/ww...t-briefs/dual-band-wireless-ac-8265-brief.pdf