iPhone 7 Plus: A Tale of Two Personalities

iPhone 7 Plus: A Tale of Two Personalities

Like clockwork, every September the entire tech world gets excited for the newest Apple device. The combination of a premium build, unmatched system performance, and tightly integrated software and services delivers what’s considered to be the gold standard in smartphone user experience.

Over the past five years, Apple’s modem supplier was Qualcomm, but this year, Apple has taken a different approach with the decision to source two instead of one baseband chipset suppliers: Qualcomm and Intel. This created two distinct RF SKUs, one limited to GSM/WCDMA/LTE support (A1778, A1784), powered by Intel’s modem, and one with GSM/CDMA/WCDMA/TD-SCDMA/LTE support (A1660, A1661) powered by Qualcomm’s modem

On the one hand, most high-end flagship devices launched this year are powered by Qualcomm’s Snapdragon 820, with X12 Category 12 integrated modem (MDM9645M discrete part). The X12 modem is capable of up to 600Mbps peak downlink speeds, 3-Way Carrier Aggregation with 256 QAM, and also 4×4 MIMO on a single component carrier. Note that it is up to the smartphone manufacturers to decide which of these powerful capabilities will be integrated into their final product.

The Verizon, Sprint and SIM Unlocked iPhone 7 and iPhone 7 Plus variants are powered by Qualcomm’s MDM9645M modem, accompanied by two transceivers: WTR3925 and WTR4905. Out of all supported features by Qualcomm’s solution, Apple has chosen to implement 3-Way Carrier Aggregation on the downlink and 2-Way Carrier Aggregation on the uplink for contiguous Band 7 or Band 41. Higher Order Modulation (DL-256QAM, UL-64QAM), and Higher Order MIMO (4×4 MIMO) have not been implemented. Therefore, the peak theoretical downlink speeds are limited to 450Mbps when aggregating three 20MHz wide LTE component carriers. We have achieved the maximum 450Mbps by aggregating 20MHz wide Band 20, Band 1, and Band 7. EVS (Enhanced Voice Services), also known as Ultra HD Voice, offers significantly improved audio quality, numerous efficiencies at the physical and IP layer, and is fully supported by Qualcomm’s modem. However, Apple has made a decision to disable this feature likely to level the playing field between the Qualcomm, and Intel variants.

On the other hand, the iPhone 7 represents Intel’s first major design win in a long time. In many ways this iPhone appears to be Intel’s make or break it in the cellular modem business. Similar to Samsung’s in house Shannon LTE modem found in Exynos based devices, Intel has decided to license CEVA DSP cores for the XMM7360 modem coupled with in-house SMARTi 5 RF Transceivers and X-PMU 736 RF Power Management. The Intel XMM7360 modem also supports 3-Way Carrier Aggregation on the downlink and 2-Way Carrier Aggregation on the uplink, but lacks support for EVS, DL-256QAM/UL-64QAM, 4×4 MIMO. Ironically, mobile operators such as T-Mobile USA and Telstra which have been offering these advanced LTE features, are being supplied with the iPhone 7 with the Intel modem.

While there haven’t been a shortage of iPhone 7 unboxing videos, subjective camera reviews, and more, we have been hard pressed to find any meaningful mention of cellular performance. This goes for any other flagship device on the market. In this day and age with mobile internet consumption at the all time high, we believe that a mobile device is only as good as its ability to seamlessly connect and maintain its connectivity with the mobile network. Most of the time mobile operators get blamed for dropped calls or session timeouts, but it’s often forgotten that the phone OEMs implementation of baseband, RF Front-End (RFFE), and the antenna design could play its role.

We have been using the Rohde & Schwarz (R&S) CMWflexx setup containing two R&S CMW500 and one R&S CMWC controller, as well as the R&S TS7124 RF Shielded Box with four Vivaldi antennas. This study has been done entirely independently, and Cellular Insights takes full responsibility for the analysis and opinions in this report. We have self-funded the procurement of iPhone 7 Plus units through commercial retail channels. All units have been preloaded with the latest version of iOS (10.0.3)

20161004_173927

Consistent with our previous reviews, our focus has been on measuring the highest achievable LTE throughput in clean channel state, starting at RSRP value of -85dBm, and incrementally reducing radiated power level while maintaining Block Error Rate (BLER) under 2%. This allows us to measure RF sensitivity of the device under test while eliminating inter-cell interference and fully controlling the radiated environment. This also allows for high reproducibility and consistency of our tests, and takes into account the performance of the entire RFFE subsystem. 

We’ve tested three unique LTE frequency bands, Band 12 (10MHz), Band 4 (20MHz), and Band 7 (20MHz) in 4×2 MIMO configuration using Transmission Mode 4. While both devices achieved the maximum sustained data rates at the cell center, simulating edge of cell scenarios by reducing power level did cause each iPhone to display two very different personalities.

As the device attaches to eNodeB, it reports its LTE capability. To get this out of the way, 4×4 MIMO and 256QAM features are not supported on the iPhone 7 Plus.

mimocapability

FDD LTE Band 12 is part of the lower 700MHz band plan, covering 15 MHz of contiguous spectrum across three blocks (A, B, C). Most LTE Band 12 deployments are either 5 MHz or 10 MHz wide even though Band 12 can theoretically be deployed up to 15 MHz widths. Coincidentally, LTE Band 12 capable devices are only certified to support up to 10 MHz operation. As opposed to mid and high band spectrum, low frequency such as 700MHz Band 12 can propagate further, penetrate the concrete structure better, and often times is the only LTE layer reaching the device. For this particular reason, high sensitivity of a smartphone receiver is extremely important in challenging signal conditions, and it could make a difference between completing and dropping a VoLTE call.

screen-shot-2016-10-20-at-2-15-37-am

Both iPhone 7 Plus variants perform similarly in ideal conditions. At -96dBm the Intel variant needed to have Transport Block Size adjusted as BLER well exceeded the 2% threshold. At -105dBm the gap widened to 20%, and at -108dBm to a whopping 75%. As a result of such a huge performance delta between the Intel and Qualcomm powered devices, we purchased another A1784 (AT&T) iPhone 7 Plus, in order to eliminate any possibility of a faulty device. The end result was virtually identical. We are hoping that this sudden dip in performance at a specific RSRP value will be further investigated by the engineering and hopefully resolved. At -121dBm, the Intel variant performed more in line with its Qualcomm counterpart. Overall, the average performance delta between the two is in the 30% range in favor of the Qualcomm variant. 

band12

Band 4 is the most commonly deployed LTE spectrum band in North America, while Band 7 deployments are spread across the rest of the globe. Mid and high spectrum bands are used to densify LTE networks and provide incremental capacity. Just like during our Band 12 tests, the iPhone 7 Plus with the Intel modem continues to struggle even at relatively higher RSRP values with unexplainable sharp dips in performance. The gap between the two variants is consistent and north of 30% again in favor of the Qualcomm variant.

band4

band7

To put this into perspective, we have compared the edge of cell performance of a few other flagship devices to see how these iPhones compare in less than favorable conditions

comparison

In all tests, the iPhone 7 Plus with the Qualcomm modem had a significant performance edge over the iPhone 7 Plus with the Intel modem. We are not sure what was the main reason behind Apple’s decision to source two different modem suppliers for the newest iPhone. Considering that the iPhone with the Qualcomm modem is being sold in China, Japan and in the United States only, we can not imagine that modem performance was a deciding factor. When all said and done, the iPhone 7 Plus is a beautifully designed smartphone, with arguably the best-in-class camera and system performance. It’s also the best iPhone ever. We hope that next year’s iPhone delivers best-in-class LTE performance. 

Milan M.P.

view all post
Leave a comment
Reviewer • 1 year ago

This is a great article. Could you please review the goole pixel cellular performance as well?

reply
brodie7838 • 1 year ago

Agreed, great article - I too would love to see this updated with the Pixel (and others!).

reply
David Bannington • 1 year ago

Could there be a connection between Intel modems and widespread reports of broken GPS/location services on some iPhone 7 and Pluses? https://discussions.apple.com/thread/7677433

reply
Jon Dagle • 1 year ago

David, the majority of problems in US seem to be on Verizon, so that's Qualcomm chip. I know of reports on AT&T and Sprint, and some European carriers--in UK I think. Many, but not all, report 10.0.3 or 10.1 fixed this.

reply
alex logvin • 1 year ago

Excellent article, Milan! Love the charts.

reply
Georgi Dimitrov • 1 year ago

It would have been an excellent one, if power measurements have been conducted too. Without it, it's nothing more than politics.

reply
Michael • 11 months ago

it is a good article without the 'power' consumption data. although that data would immensely useful as well. the fact that you sight a lack power consumption data as the article reason for not being 'good or excellent makes me suspect you of being an intel 'fan boy.'(this not meant to offensive, i am certain guilty of it myself). mainly because power consumption improvements is what intel is doing best there main line processors. with that said i would like see a power consumption test (intel's edge on smaller lithography technology is the biggest source of better efficiency ). in real world use better power efficiency/ long battery life, i would suspect to be more important to many consumers rather then the data rate performance(which they clearly prove intel's inferiority, here).

reply
Travis • 1 year ago

Yes please review the Pixel radio compared to other phones. Out of all the reviews I have seen online for the Pixel no one actually reviewed the "phone". After all we demand these devices work well when no WiFi is around.

reply
Thomas Anderson • 1 year ago

Please do not review Google Pixel. I don't care the result of it. Thank you.

reply
Dazza • 1 year ago

Please dont read the article on the Pixel. Thank you

reply
John • 1 year ago

Any way to determine which variant is in my personal iPhone 7 plus?

reply
Steven H. • 1 year ago

Yes, easily. The article noted which carriers are selling the one with the Qualcomm modem. If you bought it outright, was it for use on one of those carriers?

reply
Holger • 1 year ago

The model number of your iPhone is printed on the back of the device.

reply
Brian B. • 1 year ago

You can also find the model number by going to Settings > General > Regulatory.

reply
ilikemyprivacy • 1 year ago

A few thoughts: 1. Thanks for doing this, so little information available on RF performance of phones which is so critical for performance, battery life, etc. 2. I would be interested to see you rerun this test with the iOS 10 Beta. The latest beta includes modem firmware 1.25.00 on the Qualcomm version which seems to help significantly with a variety of issues 3. If possible, it'd be great to look at a comparison for other RF front end features like envelope tracking, antenna tuning, etc. between the Intel and Qualcomm models as we..

reply
Walt French • 1 year ago

Great article; thanks. Maybe I missed it, but I didn't see in the article what features in the new Intel modem differed from the 6S's MDM9635M. As the comparison chart makes the two parts look very similar, what appears is that the 9645's new features are (today) a year ahead of the 9635 and the Intel part.

reply
Marcus • 1 year ago

My question is the one with the Qualacom modem vs Intel was there any correlation between battery life too? If the qualacom works better does it use more battery power based on tests than the Intel. Just trying to gage it if i do decide to purchase a iPhone 7

reply
Flash • 1 year ago

Thanks for sharing. Was this test conducted using the latest update iOS 10.0.3 which includes a fix for a bug that could cause issues with cellular connectivity, a problem that only affected the new iPhone 7 and the iPhone 7 Plus? Thanks.

reply
Dazza • 1 year ago

Yes. It says so in the article

reply
Edvinas Maciulis • 8 months ago

its clearly written " All units have been preloaded with the latest version of iOS (10.0.3)"

reply
Tom • 1 year ago

Great article, I really liked your introduction paragraph!

reply
Maxwell • 1 year ago

Can you do a review on the ZTE Axon 7 signal strength and LTE performance please and thank you

reply
Prasad • 1 year ago

Marcus, Assume you are downloading a file that is 70MB in size. If your iPhone with Qualcomm chip downloads this in half the time, it will be able to go to sleep that much faster. It is not necessary that the performance is obtained at the cost of power.

reply
VP • 1 year ago

So if this is the case my iPhone 7 Plus purchased thru T-Mobile should not have any Verizon or Sprint connectivity. However when accessing carrier in settings it scans for available networks & finds T-Mobile, AT&T, Verizon & Sprint. It should not be possible due to hardware limitations. I couldn't get mine on release day even with pre-order. Could the demand & backlog have prompted Apple to add Qualcomm modems to all models?

reply
WiWavelength • 1 year ago

VP, the available network scan on your 3GPP only baseband is detecting the LTE networks for VZW and Sprint, not their CDMA2000 networks. The latter use PRL based network selection, not user initiated network selection. AJ

reply
Eric • 1 year ago

It would be great if you could provide a picture of the test setup.

reply
Yang • 1 year ago

Without testing environment and setup specified, this type of comparison is less meaningful

reply
Jupiter Antunes • 1 year ago

One important point must be kept in the picture, not only the modems makes the differences but the Antenna with the Modems. One significant information should be the Return Loss of the antenna scanning the channels in a Band. If a LAB is reading this could do it for us... thanks.

reply
Ehsan • 1 year ago

Qualcomm should ask Apple for "Qualcomm Inside" sticker on the phones with Q chip.

reply
no • 1 year ago

no

reply
Michael • 11 months ago

yes

reply
Joe • 1 year ago

If you swapped an Intel variant iPhone out at the Apple Store under warranty, would they supply you with another Intel or do they use the Qualcomm version for warranty replacements?

reply
Justin • 1 year ago

Whenever you turn in a warranty claim, most likely you will receive a refurbished unit. Not a brand new one when you first buy/order. I would return it under the 30 day window and purchase another one personally. Although the phone is new, you could still potentially get a refurb

reply
Robert • 1 year ago

It wold be interesting to see how the two different transceiver (RF + modem) compare in case of carrier aggregation that are in harmonic relationship, such as B12 as the primary carrier transmitting at 710 MHz and B4 as the secondary carrier with receive frequency at 2130. In this case the 3rd harmonic of the B12 transmitter lands right in B4 receiver thus potentially de-sensing it and degrading the throughput when the B12 TX power is high (cell-edge scenario).

reply
Ed • 1 year ago

I am still waiting for Anandtech to do their digging. And offering some explanation to why. The Intel Modem is based on TSMC 28nm. Qualcomm X12 is on 20nm. ( Only the X16 is on 14nm FinET I believe ). X12 is / could be even better then the coming Intel offering 7480, which is STILL manufactured by TSMC. Why did Apple decide to use the newer X12 when they know the older MDM9635 was enough compared to Intel's 7360? Not supporting 4x4 MIMO in X12 was a rather simple decision because less then 10% of the network in the world support it, and no phone has yet to manage 4x4 MIMO inside. A some other details like X12 wasn't a true 4x4 MIMO and require separate chip to do so, while X16 is a true 4x4 MIMO. And I am pretty sure an Software update could make EVS possible. Whether or not they will do it is entirely different matter. What cost / patent issues are involved?

reply
Luka • 1 year ago

How it could be achived more than 150M for SGS7edge on one carrier Band4?

reply
PD • 1 year ago

4X4 MIMO

reply
R • 1 year ago

2X2 MiMo single carrier using 64QAM is 75Mbps x 2... 150Mbps and 3CA with same config is 150Mbps x 3 giving 450Mbps peak.... too bad Apple did not use 256QAM!

reply
Luka • 1 year ago

Then I would expect more than 180Mbps

reply
Bjoern • 1 year ago

Nice rewiew, and measurements on the poewer drawn by both modules when performing differently?

reply
Mark • 1 year ago

This is a great article. I happen to live in an area that gets marginal reception but I never knew it because I had a Nokia 928 from Verizon. When i finally had to trade it in after a couple of years, I tried the iphone 6S and found the reception to be horrible. I traded it in for a Galaxy S6 and was told "no more trades" so I was forced to keep it. That also had horrible reception at my house. My signal varies from -105 to -115 and calls constantly get dropped. My girlfriends Galaxy S5 worked great at my house and I took that as a sign as the antenna design changing in the Galaxies between the S5 and the S6...which I found to be true. When her phone was ready for replacement I searched around and found no bad things were said about the HTC10 reception. At my encouragement, she got the HTC 10 and gets great reception at my house. I have been searching for some sort of articles on antenna testing and this report is just what I was looking for. I would, however, like to see a wider range of phones tested. It would be interesting to see how that HTC10 does against the iphone and galaxy. And, I would also like to see how the new Droid Z and the Pixel fare. thank you again

reply
Alexander • 1 year ago

Thank you for your effort, Milan. It would be nice to see the error bars on the graphs and the statistical comparison for the low signal LTE examples. Given the low average current networks speed limits (9-40 Gbs) it would be crucial to magnify graphs accordingly. I am going to use my phone mostly on the European GSM networks, so having A1661 with CDMA capability would not grant automatic CDMA network switch with European sim or ...?And of course after all it would be nice to prove the efficiency of your 30% discrepancy in real world or close to real world situation...What would be battery impact?

reply
John • 1 year ago

The disabling of HD Voice - is it permanently disabled or is that a function that Apple could enable in the future?

reply
Adam • 1 year ago

Lab test isn't field test. If you do field test,the results are almost same.

reply
Darrell Brand • 1 year ago

Just called Apple and mentioned my concerns- I bought my iPhone 7 unlocked and no plan Apple has no response.

reply
landoncube • 1 year ago

The SIM unlocked version is the Qualcomm modem, according to the article.

reply
JohnL • 1 year ago

Only Qualcomm in the US and a couple of other places, it's Intel in Europe (1778 model is supplied as the unlocked one).

reply
JohnL • 1 year ago

What was the modem firmware used in the tests please? Also IMHO worth repeating in 3-4 months to see if Intel have made significant improvements to a young product..

reply
JohnL • 1 year ago

I note 10.1 updates the modem firmware (Intel model) from 10.0.3's version 1.00.05 to 1.01..13, which seems like quite a bit. I don't know what was in 10.0.2 I'm afraid.

reply
JohnL • 1 year ago

10.1.1 takes it to 1.1.15

reply
JohnL • 11 months ago

We're now up to modem firmware 1.02.15 and the performance is definitely getting better,

reply
Jacq • 1 year ago

This article made for great reading. A couple of things I would be interested in seeing in the future articles. 1. Testing this on Google's Pixel device along with every major flagship release. This is information that would influence a purchase decision. I also from daily use between the 5x and 6p felt the 5x was horrible and which you tested the 6p's signal as well, 2, Was there a measure of power consumption at the different signal strength between the two chipsets? Wondering if this affected the performance as well. Thanks - JSH

reply
Alessia Campoli • 1 year ago

Good to see that iPhone 7 will be IP67 waterproof! When concerning about the waterproof function of mobile, I found an very interesting video about it. On this video, they drop VKworld T1 Plus, a budget 6.0-inch smartphone to the mud and clean it directly with water for long time, that is unbelivalble.

reply
VG • 1 year ago

Even if you leave apart the intel vs qualcomm modem differences, the article clearly tells that SD 820 based devices will have far better connectivity as compared to any iPhone7

reply
AP • 1 year ago

If I'm on the AT&T network, is there anyway to check if I'm getting a Qualcomm version before I purchase/open the box?

reply
Alexis • 1 year ago

the model number is written on the outside of the box. You could read it without breaking the seal or opening the box. The ATT model uses the Intel modem. If you want the qualcomm then buy a unlocked version

reply
landoncube • 1 year ago

My research indicates that there is no AT&T Qualcomm phone available. I am looking into the SIM unlocked Qualcomm version for use on AT&T.

reply
Doug Lerner • 1 year ago

I live in Tokyo and use KDDI Au. I can't find out which model they are using. AU shops say "it might be either, but you can't tell until purchasing and opening the box to look." All the floor models I've seen have the model number with the Intel chip though. I'm wondering how much of an issue this is and if I should upgrade my iPhone 6 Plus now.

reply
karyn • 1 year ago

the model number is written on the outside of the box. You could read it without breaking the seal or opening the box.

reply
Doug Lerner • 1 year ago

You're right! I can see that on the back of my iPhone 6 Plus box. That precludes being able to upgrade online, but perhaps I can do that in a store. Assuming there is even more than one model sold here in Japan for AU. I still haven't been able to confirm that yet.

reply
Kari • 1 year ago

Let's wait a little and see what's Apple's response is, before jumping into conclusions.

reply
Doug Lerner • 1 year ago

Actually, it turns out that the model sold here in Japan is neither the A1784 nor the A1661. For the iPhone 7 Plus it is the A1785. The details are at http://www.apple.com/jp/iphone-7/specs/ but the modem supports the following: FDD-LTE(バンド1、2、3、4、5、7、8、11、12、13、17、18、19、20、21、25、26、27、28、29、30) TD-LTE(バンド38、39、40、41) TD-SCDMA 1,900(F)、2,000(A) CDMA EV-DO Rev. A(800、1,900、2,100MHz) UMTS/HSPA+/DC-HSDPA(850、900、1,700/2,100、1,900、2,100MHz) GSM/EDGE(850、900、1,800、1,900MHz) I don't know if that means it's Qualcomm or Intel. The Japan model also supports FeliCa, which allows people to use it like Suica/Pasmo cards to ride the trains and subways.

reply
Pei • 1 year ago

it's a qualcomm variant, not intel.

reply
Pei • 1 year ago

this article has been seen in China. Only Qualcomm variants are sold here. Many apple users are not satisfied with APPLE because iPhone is expensive while Apple used a bad modem.

reply
JohnD • 1 year ago

All devices with qualcomm have issues in the field. Lab vs field tests are completely different. The mobility tells how feasible a device is. Not speaking about the fact that in real NW the eNB will never give yous such a bandwidth. I really don't understand why you people struggle to have in the pocket the newest devices on the market...

reply
ElmCha • 1 year ago

All Qualcomm devices have problems in the field? How so? Kindly elaborate.

reply
Michael • 11 months ago

+ elaborate i am not that familiar cell antenna, but article does appear to show Qualcomm having substantial edge of intel here.... with that said i avoid computers /network devices and other kinds of 'smart' devices that utilize qualcomm networking chipsets like the plague, when i have choice anyways. (performance is not usual my concern, its their buggy and incompatible nature that i dislike the most )

reply
Alexhander • 1 year ago

Great article! Very nicely detailed and accurate information based on the currently released iPhones.

reply
salman Khan • 1 year ago

Both intel and Qualcomm modems are capable of supporting more than 400Mbps.if LTE network give them enough resources. and channel conditions are good. This means that performance of the intel modem may not be limitation here. It depends on test conditions, transport protocol parameters, both mobiles (Qualcomm and Intel) exact locations and many other factors. As a user, I dont care much if my downloading speed is 150Mpbs or 140Mpbs because I rarely run any application on my mobile phone that requires this much high data rate on my mobile(HD video stream needs few Mpbs speed). But I would be more interested in power comparison and battery life with these two modems.

reply
Bob H • 1 year ago

My burning question is whether the inferior Intel modem I would get in my AT&T 7 Plus would also have inferior receptivity to my AT&T iPhone 6. If so, I will wait to upgrade until the 8. I already have spotty reception at home with my 6 and I can't afford to make it worse. Is there already a way to compare this or does it require a new test? Thanks.

reply
Bob H • 1 year ago

I posted a question here that was held up for curation and then disappeared. It seemed perfectly harmless. What happened?

reply
Bob H • 1 year ago

I had a chance to answer my question by doing a field test on my 6 at home and learning I was getting a reading of -114 in most places, and then bringing home a 7 Plus (with the inferior Intel modem) and getting readings between -108 and -112. Glad to learn that "inferior" doesn't mean worse than a 6, in my location at least, so I'll keep the 7 Plus.

reply
landoncube • 1 year ago

Could you repeat the testing now that iOS 10.1 has been released? I have an A1784 iPhone 7 plus.

reply
JohnL • 1 year ago

It is depressing how much my 1778 is slower than my 5s for LTE. A quick 4G test (or few) this evening gives 65.7MB/s down/14.1 up for the 5s and 60.0/6.5 for the 7... Plus I only tested it as I'd seen a trend so this wasn't a one-off. A bigger concern is the 5s holds onto a 4G signal much better, the 7 does like to be on 3G (iOS 10.1).

reply
Jos • 1 year ago

There is a very interesting article about phone reception quality. Its writen by professor Gert Frølund Pedersen, from Denmark's Aalborg University. The direct link to the PDF: http://www.pts.se/upload/Rapporter/Tele/2016/MobilephoneTest2016-augusti-2016.pdf I ran some tests myself on the 3 Belgian providers. Reception quality inside my house is very bad where I live (bad coverage) with signal levels between -95 and -113 dBm. By the way, I'm an engineer and have quite some RF knowledge. Tested with a brand new iPhone 7 (Intel modem), iPhone SE, Nexus 5, Lumia 925 and Lumia 800 (both Lumia's are still Nokia design, not Microsoft). Tests are pure subjective, on different places in my house with all the phones and swapping SIM cards every time. With a signal of -95 dBm, the iPhone 7 often doesn't receive or can't make calls and text messages can't be send. More or less the same story with the iPhone SE. The Lumia's perform the best, the Nexus 5 sits between them. The Lumia's always have a better signal then the other phones. I don't care about LTE performance (I only get 3G at best over here) since I have Wifi in house. Don't understand why reviewers talk about all the gadgets of the phones, while the basic functionality (making phone calls) isn't even decently tested in an area with bad reception. Of course it's much more interesting to compare stuff like camera functionality, screen colours, quality of the Gorilla glass, bending phones etc, but it's all non-essential. I'd love to have a phone that bends when you sit on it, but at least has a good reception and can be used to make a call. Just want a phone with the best receiver sensivity. For each 3 dB decrease in level, the singal power is reduced by about one half. So a more sensitive phone just works better in environments with weak signals. That also includes using it in your car, a bad environment for a mobile phone (metal cage). And most of the time, the receiver in the phone is not the problem, but the antenna plays a much bigger role... As professor Pedersen suggests in his article, it becomes time manufacturers are open about this in the specs of the phone.

reply
Massi • 1 year ago

What process node is the Intel modem? Would be very interesting to see power comparison between Intel and Qualcomm

reply
Psymon • 1 year ago

Great article.. I found it just in time! Having just purchased an iphone 7 plus, 32GB on contract in the UK for use on the Three network. Without any hesitation I have now returned the device and will wait until the i8 appears. Here's why... I live in a network fringe area of Cornwall and also trunk travel quite a bit. The fringe signal at home for the i7 displayed a significant decrease on the previous i5 and i5s signals. 4G has been incredibly sporadic everywhere had appeared to offer no performance increase over 3g. Another immediate issue was a very sharp spike in data usage, I have been using the Three, All You Can Eat Data plans for years, peaking at 25gb in the heaviest usage months... In just five operational days of the i7 the daily use started to top 3gb when streaming music, using google maps and general browsing, close to 50% more than I had ever seen before. I spoke with Three technical for a few hours, explaining everything in great detail, in an attempt to overcome the issues or at least gain a good explanation. Three almost immediately admitted that the 4g was not working on the i7 properly and a solution was being sought between Apple and Three (Apple deny this) Three technical suggested turning off the 4g until being advised of an update that resolves the issue. There was no explanation as to why the signal was poorer on the i7 compared to all previous devices, but now I am starting to understand why, your article makes perfect sense. The Data usage increase was also not met with a good explanation but was dismissed by pointing out that I already have All You Can Eat Data, so why worry.. point taken... but be aware if you are paying for capped data limits! By the way the upgrade to iOS 10.1.1 has made no observable difference.! I also ran a performance test on the i7 using a little app "Performance Test" just to see whether the phone performed as advertised.. A friend also purchased an i7 with 256gb at the same time as me giving me the opportunity to compare both. What was remarkably disappointing was the write speeds... the read speeds were very similar but the write speed was significantly slower on my 32gb i7. Close to 50% less capable slowing the device down by 50% when writing to its memory. This was also observable when browsing images on both phone with maximum signal strength. (My friend has also retuned his iPhone 7 for the same reasons) As a final pass before returning the i7, with your article in hand, I made an appointment with Apple to see if they were prepared to shine some light on the issues. To be fair, they were not really interested, they did however offer me a replacement but with the Intel Model numbers. They confirmed that the Models with the Qualcomm component installed were not available in the UK. If you have an iPhone 7 with the model numbers A1778 or A1784 or are considering buying a new phone.. DON'T it really isn't worth the grief and they will be virtually worthless once the issues are widely known.. These phones WILL give you trouble, perhaps not so much in the city but rural and travelling they simply do not meet a price tag of £50.00 for a phone let alone £780.00 , they really are no more than just a very expensive camera/ipods.

reply
Peter • 1 year ago

Three days ago in Munich I habe bought an i7+ delivery mid Dec. After that article I'll cancel the order and keep my 5se waiting for i8. GREAT test review - a must read for any Apple user.

reply
Peter • 1 year ago

Upon canceling the order I've asked myself what a mobile phone is for - if not primarily for mobile connection like talking and data transfer, Otherwise it's called iPod. I7 is close to be seen as an iPod with the occasional option of mobile connectivity. Whatever was presented at the last keynotes - it was never any comment on "we offer great mobile connectivity" . It was an unexpressed and intrinsic property of any iPhone as we all have assumed. This seems to be wrong in the post Steve time since the i5 was the last one approved by him. Doubling profit and revenue the basic feature of the iPhone went lost: talking. Wasn't there any product manager responsible for? Others have their battery gate but a mobility gate is even worse. Iwonder what Woz' opinion would be. I would love to read a statement from Apple. And: this testing and benchmarking shall become a standard from now on watching the i8 and others.

reply
Peter • 1 year ago

These values do match the findings. https://www.t-mobile.de/sar-werte/0,17595,1347-_,00.html The weaker the mobile efficiency the higher the power to compensate and so the impact to brain. The explanation is found here https://www.google.de/amp/appleinsider.com/articles/16/11/18/apple-confirmed-limiting-iphone-7-qualcomm-modem-to-keep-performance-on-par-with-intel-chip/amp/

reply
Peter • 1 year ago

Conclusions: Apple uses 2nd sourcing to keep purchasing price low. Regardless performance or radiation impact. There was a time Apple was standing for perfect products. Not anymore. Only iOS is binding the customers finally to Apple.

reply
matt • 1 year ago

I am so glad I found this article on the I phone 7 plus, living in the uk and on 3 network I just discovered after upgrading to the iPhone 7 plus this issue with signal and how poor the intel version of the phone is and its reception. living in a marginal reception area I have decided to cancel my upgrade and go for something else, its nice to have a phone with unlimited data but only if you can use it reliably. mostly I want a good signal over anything else and from what I have read in these last few hours the iPhone 7 plus suffers badly with poor reception on 4G here in the uk. Apple should be ashamed of the iPhone 7 plus and its reception performance. for a high end product at high end price. I would want it to out perform its predecessors and be an improvement not a step backwards in its main function as a phone that needs a signal. very poor R&D Apple and you should be ashamed to not even come forward and admit to your loyal customers you made a mistake with the 7 and 7 plus. cancelling in the morning. Thank you for article.

reply
Chris • 12 months ago

I appreciate the detail in this article. I started searching for info like this because the rest of my family has iPhone 6s Plus's. I now have a IPhone 7 Plus. To my disappointment, their phones outperform my new phone when we play Pokemon Go together. They start to catch their spawned Pokemon before I even see mine. They laugh at me and my new phone. I also lose connectivity with the Niantic servers more often than they do. When I called or chat with Apple Support, they just say, "It's the software." All the phones mentioned, are on the same AT&T account (8 total phones). When I went to the AT&T store, they said they don't deal with performance issues. I had to contact Apple. When I call Apple, they say, "It's the Carrier or Software company. This has been very frustrating.

reply
ebony threesome • 9 months ago

Great blog post.Really looking forward to read more. Really Great.

reply
SMS • 9 months ago

So, what is the so good about this phone anyway? it incredibly expensive, inferior camera that is producing low quality images, buggy OS, and now, prone to failures performing a basic tasks a simple phone should be able to do, like sending and SMS, and now this... I wish I kept my older phone, I really want to give apple a piece of my mind, like face to face.

reply
Shayna • 8 months ago

I have the iphone7 A1778 Intel model with T-mobil. It constantly cuts out in the middle of phone calls where I can only hear the caller, however they cannot hear me. Everytime I've spoke with representatives, they initiate a 'software update' or a reset process etc, and then make calls to themselves on their T-mobile network- which go through fine. I only have the problem- it appears, when calling non-Tmobile phones. I have figured out that I can temporarily rectify this issue by placing the call on speaker mode for a few seconds and then returning and the caller can hear me again. However I end up doing this several times throughout the phone call. Is this a known issue for this particular model? Have there been any known work arounds? I'm just past the 30 day return, and it seems there is a huge issue with this phone! I'm very upset that I just now am reading this information on the different modems- just a bit too late for my own good!

reply
George Mavrikas • 7 months ago

How about revisiting this article, now that apple has updated the 7 a few times. Maybe they have fixed the issue or narrowed the diff.

reply
Jonathan • 3 months ago

Can we revisit? Lots of OS and firmware updates since this article was first published. Intel firmware seems to be on version 2.0.0.1 now (iOS 10.0.3 used firmware version 1.00.05). It would be very interesting to see if performance has gotten better over time. It would also be nice to include iPhone 8 as it appears Apple uses both Qualcomm and Intel modems again. I think i'd be interesting to compare both 7s and both 8s to see whether the gap has closed, if there was improved performance via firmware updates, and if there was improved performance with the next generation.

reply
Nooxm • 2 months ago

Can you please do a review ab iPhone 8 and iPhone 8 Plus?

reply

Leave a Reply to Bob H or Cancel Reply

Please be polite. We appreciate that.

By Daniele Zedda • 18 February

← PREV POST

By Daniele Zedda • 18 February

NEXT POST → 34
Share on