But if we want to assume …
… then https://www.qualcomm.com/media/documents/files/snapdragon-632-mobile-platform-product-brief.pdf includes “Integrated 802.11ac multi-user MIMO (MU-MIMO)”.
Edit: Ah, I see, it’s not mentioned in https://www.qualcomm.com/products/snapdragon-632-mobile-platform … quite confusing.
Ha ha! I’m out of my depth here. Can’t think of a single use case where I needed high bandwidth in the past. So probably not remotely important for me. I hesitate to ever use the term WiFi 5 again! 
My portable external monitor didn’t arrive until after I’d gone and broken the screen on my fancy smartphone, so I never tried using the smartphone like a desktop. I can imagine if one were to do so though in Android 10 desktop mode, with multiple browser tabs open and on the go, then bandwidth might possibly be of some importance. I now know about Emby, Kodi and Plex though. Is NetFlix even available in 1080p on Android (don’t bother answering this question)? Don’t use NetFlix any more anyway.
Basically, I have no facts to add.
Yes, Qualcomm contradicts itself on these 2 first-hand sources.
The FP3 will have 802.11ac (and Bluetooth 5.0) according to the earlier linked tests.
I’m not sure what to make of it. One source supports the 632 narrative, the other one does not. We don’t know which one is accurate, do we?
I got more bad news for the 632 narrative and/or 802.11ac proponents. We can look at other 632 devices.
The Redmi 7 is the device from Xiaomi which sports a 632. According to Xiaomi’s website it sports 802.11 b/g/n and Bluetooth 4.2.
https://www.mi.com/redmi7/specs/
无线网络
支持协议 802.11 b/g/n
支持 2.4G WiFi / 支持 WiFi Direct / WiFi Display
支持 蓝牙4.2
支持 收音机
Which translates to
wireless network
Support Agreement 802.11 b/g/n
Support 2.4G WiFi / Support WiFi Direct / WiFi Display
Support Bluetooth 4.2
Support radio
Wi-Fi.org source:
http://certifications.prod.wi-fi.org/pdf/certificate/public/download?cid=WFA82628
I’m smacking my head to the wall with regards to Wi-Fi.org’s search engine. I can’t find the Motorola Moto G7 Power. According to
https://www.gsmarena.com/motorola_moto_g7_power-9527.php
It supports
WLAN Wi-Fi 802.11 b/g/n, hotspot
Bluetooth 4.2, A2DP, LE
Same as the Redmi 7, and same SoC.
EDIT: Comparison between the 632 and 636 is here. It does not mention WLAN. https://mygadgetreviewer.com/snapdragon-632-vs-636-comparison/
EDIT: If its possible that Motorola and Xiaomi simply did not use 5 GHz antennas, to avoid size/cost/price differentiation/etc then that would solve the mystery. It would mean the SoC in these devices (632) is capable of 5 GHz but they decided to omit the further hardware support for it.
EDIT: Honor 8c sporting the 632, also only has 802.11b/g/n 2.4 GHz https://www.gsmarena.com/honor_8c-9366.php
Whatever the the exact name of the chipset, what matters is real world performance. I switched from an FP2 to a (refurbished) Pixel (2016) last year and was amazed by the performance. Even today that device is a joy to use and comparing the Geekbench score, it seems that the FP3 will be comparable in single core performance and around 25% faster in multi-core performance. Most metrics look really well compared to the Pixel, expecially importand tasks like parsing and handling webpages (HTML parsing and dom performance). AES (encryption) performance is really important as well, since the FP3 will mostly likely be encrypted by default.
Find the benchmarks here:
Device Single Core Multi Core HTML AES
Fairphone 953 2498 3240 190
Pixel 2016 1643 3492 5252 2073
Fairphone 3 1275 4902 8698 3317
Looks like the used chipset was a good choice.
For me 2.4 Ghz is completely unusable because that band is overcrowded where I live. I deactivated it on my router. Unfortunately I have two devices that can only ‘speak’ 2.4 Ghz: my printer and my ebook reader - I have to feed them via USB.
Both bands can be served with the same antennas in parallel.
Why do we think it’s a Qualcomm chipset?
@JeroenH for Motorola you need to search for the product code which for the g7 power is xt1955 I think, but please confirm.
Because Geekbench mentions “Qualcomm Qualcomm” as the processor name. See the link in the second post in this thread. Additionally that page states “81” as the implementer. That’s the decimal value for the capital “Q” character, short for Qualcomm. Might sound speculative, but as someone who briefly worked for ARM I happen to be able to confirm that that’s exactly how ARM has chosen implementer codes for the few parties that license the ISA to implement their own cores rather than just use ARM-implemented designs.
Yeah that’s it, cheers.
So, I started with the top and -1 -2 -4 -7 only have 2.4 GHz. Interestingly though the variations -5 -6 -DL do have 5 GHz, though in form of 802.11n. None of them appear to have 802.11ac.
Did you mean the Fairphone certification?
https://www.wi-fi.org/product-finder-results?sort_by=model&sort_order=asc&categories=4&companies=9084278
It’s WiFi 5
Direct link to certificate
And it’s more proof to me, that the phone will be ready for delivery upon presentation.
Geekbench is using the same term for the Xiaomi Mi A2
https://browser.geekbench.com/v4/cpu/compare/8635861?baseline=8477654
For motherboard it states: sdm660
The same goes for the Nokia 7 plus
https://browser.geekbench.com/v4/cpu/14332601
And both times it’s base frequency 1.84 GHz.
So, maybe it’s a Snapdragon 660 ?
See this article on the Motorola G7:
https://www.mysmartprice.com/gear/motorola-g7-3gb-ram-variant-geekbench/
The benchmarks (1260 single core and 4759 multi core) is very much comparable to the FP3.
I guess, I now would take you up on it.
Just kidding of course. 
The 660 I found systematically had “Qualcomm Qualcomm @ 1.84 GHz
1 processor, 8 cores”. FP3 and other 632 systematically has “Qualcomm Qualcomm @ 1.80 GHz 1 processor, 8 cores”.
The 660 I found had far better singlecore and multicore performance than FP3. FP3’s single- and multicore performance was aligned with 632 performance.
Who submitted these Geekbench entries anyway? I suppose Fairphone employees. Can they be spoofed?
100 EUR. No joke 
As for what you added afterwards,
https://www.gsmarena.com/motorola_moto_g7-9357.php
is a 632. Yes, 632 have similar performance as FP3.
I was comparing with the Motorola G7 Power, so I was looking for that one.
I already discussed and showed with screenshot from the PDF, that FP3 has WiFi 5:
EDIT: Also keep in mind (to add to my written point #1 and #2) that the memory performance of the FP3 is lower than the 660. From my PoV (layman) that may suggest LPDDR3.
(Yes sorry I modified my post a few times, bad habit to post and edit…)
No, I never bet.
And I give in to superior knowledge and research. 
I just fell for “fake news”, as the article states (and my wishful thinking):
The “ARM Implementer 81 architecture 8 variant 10 part 2048 revision 2” identifier, however, makes it quite clear that it is actually the Snapdragon 660.
And I have to admit as well, that I had missed your posting of the WiFi certificate and did not read the post you answered to thoroughly.
Sorry for that.
My impression:
From the geekbench numbers, the SD636 is the chip.
It scores about 4900 in multicore.
Fairphone 3?:
SP636 Octa core.
Supports 18-9 display.
Dual sim.
24mp(only one camera on picture) or 2x16mp camera.
Biometric scanner on back.
Bluetooth 5
Quick charge 4.
LTE with 600mbs.
Battery = any inputs here??
Think by logic USB-C is the only logical solution. EU have made it standard port for smartphones?
What do you think?
EDIT: Also from the open fp3 on the desk(picture) , it looks like 10 or 12 screw from disassemble of the fp3, the fp2 had 9 screws. So about the same modules?
To the right is also a pair of mini jack headphones.!
No worries, I just put in some time for it partly during “work” I’m no processor or mobile or embedded specialist like e.g. RSpliet who worked at ARM.
I’d also prefer a 660, but TANSTAAFL. It comes with a price. I think the performance of the 632, from what I’ve gathered, will give me a phone which lasts 4 years. (As long as my kid doesn’t use it as boomerang.)
I must have been blindfolded, at least partly. The SD 632 actually has four Kryo Gold (i. e. out-of-order) cores. It’s only disadvantages are USB 3.0, a weaker GPU, a slower modem and slower WiFi. I could easily live with that. 
Just to clarify one thing, that is also found in the spec sheet I linked earlier: Qualcomm allows phone manufacturers to pick and match certain features in the SoC based on the price they wish to pay for the SoC. Consider some features (like .ac WIFI or the set of supported LTE frequencies) optional extras the same way you can pick optional extras when you buy a car. Generally features are enabled/disabled by means of dedicated firmware or through one-time fuses that Qualcomm pops to disable features the customer hasn’t paid for.
Why I bring this up is because it is another reason why different phones with equal SoCs could still have different specs. Not finding a phone with a Snapdragon 632 and 802.11ac wifi doesn’t mean the SoC can’t do it, but perhaps manufacturers just didn’t want to pay for the feature because few consumers really ask for it. Seeing a certification for the Fairphone 3 doing LTE in “category 7” is a strong hint that they will use SDM 632 SoCs in the next phone, but it could also still be a 636 for which FP didn’t pay the license fee required for cat 12 support.
Until Fairphone confirms, there’s insufficient data out to give any guarantees on the exact SoC… but the WIFI and LTE certificates do give valuable information on the final specs of the phone, which in the end matter more than the version number stamped on the SoC.
The processor information on geekbench labeled “identifier” rather seems not to identify the processor.
I have found the string: “ARM implementer 81 architecture 8 variant 10 part 2048 revision 2” for quite a few phones (the search shows “revision 1” as well though) with different processors, like:
Xiaomi Mi A3 with a snapdragon 665
Xiaomi Mi A2 with a snapdragon 660
Nokia 6.1 Plus with a snapdragon 636
Motorola Moto g(7) with a snapdragon 632
See …
all the same identifier but quite different identities
Even the given MHz values are differing between 1612 / 1804 / 1843.
Whatever is identified by this term, it’s not the processor as such. At least it seems to be a qualcomm snapdragon 6xx all the time.
I excluded everything which didn’t fit 1.8 GHz because the 632 I saw with 1.8 GHz had very similar performance.
The Xiaomi Mi A2 and A3 have better memory bandwidth. The A3 furthermore, has better performance as well. 660 is out of the question.
The Motorola Moto G7 you linked has very similar performance as the FP3 which further supports a 632 narrative.
I did not mean to imply anything else.
I am actually really convinced already.
I just wanted to show, that this term does not help to clarify anything.
And maybe someone has any idea, what this identifier on geekbench really identifies?