Starfyre's T-Mobile/AT&T (Intel) vs Verizon (Qualcomm) iPhone X Benchmarks for GSM Users

[Starfyre]

So after reading a lot of articles from reddit posts and articles to forum posts claiming how if your on T-Mobile and AT&T, getting the Verizon model is better because Qualcomm is better, faster, better battery life, etc. I wanted to see for myself how the two compare and determine whether the Verizon or T-Mobile one is better. Hope this helps everyone out with the exact same question and issue as me! Conclusion after the pretty graphs, more about the setup of the experiment and aggregated data/averages below that. (as an aside, I think this is the longest starting post in any thread I've ever created)

If anyone else has tests and observations they would like to share that you notice from these results, please share!

Pretty Results Up Top (Graphs), 11 Rounds Total, names of each round on left side

Download Summary: Qualcomm beats out T-Mobile/Intel on 8 rounds. Intel wins 3 rounds.
Download Winner: Qualcomm

Ping Summary: Qualcomm beats out T-Mobile/Intel on 7 rounds. Intel wins 3 rounds. Qualcomm is tied with Intel in one round.
Ping Winner: Qualcomm

Upload Speed Summary: Intel wins 8 rounds. Qualcomm beats out T-Mobile/Intel on 3 rounds.
Upload Speed Winner: Intel

Conclusion
I'm leaning towards keeping the Verizon/Qualcomm iPhone X and return the T-Mobile iPhone X because:

1) I think Download Speed and Ping are more important than the Upload Speed.
2) The max and min download speeds (though I don't explicitly include in the post of the Qualcomm when it wins are almost always higher than the min and max of the Intel.
3) T-Mobile dropped cellular data/internet connection in the middle of testing one of the cases and pulling up a web browser didn't work either. Connecting to Wifi, then disabling Wifi did not work either, but after taking the sim out and putting it back in after a minute or two, it started working again. The Verizon did not have such connection issues running through the 11 rounds (each round with 3-6 trials, mostly 5-6 for the majority) I had put it through. Even though it happened once on the Intel, I can confidently say from personal testing, the Qualcomm is more stable, even though it happened on the Intel only one time throughout my tests.

Though I didn't capture any data on battery life, from my observations, after running through the morning rounds, battery life appeared to remain at a 5% decrease on both devices, and measuring some battery life at the end of each round, the battery life appeared to decrease at about the same rate. Others can probably perform some further studies, but I have not been able to conclude that one is more power hungry than the other (at least not to any noticeable degree).

What do you think? Do you think otherwise based on the results? Do high upload speeds matter to you?

Tools Used
iPhone X T-Mobile
iPhone X Verizon
T-Mobile Sim (same Sim and Service used for each and every test)

SpeedTest Server Locations
I used 3 different servers, for the purposes of being anonymous, I am not going to specify which ones they are, but they have the following characteristics:

Server A: Closest to my location in the same state and city
Server B: Slightly close to my location in a neighboring state
Server C: Farthest from my location in another state

Rounds / Times of Test
Each comparison between Verizon/Qualcomm vs T-Mobile/Intel on a single server is a Round. Each Round is an average of 3-6 trial runs at 3 different times in the day (Morning, Afternoon, and Evening). While one phone is being tested, the other is in standby with display off.

Setting Up Experiment

Setup
- Started up both phones, installed Ookla SpeedTest app.
- Auto-Brightness Disabled in iOS 11
- Disable Wifi
- Shut down both phones
- Start up both phones
- Launch Ookla SpeedTest app

Runs
- I manually select Server A, B, or C
- Hit the start/run test button
- Repeat n times for each server
- Whenever I start a new time “Morning/Afternoon/Night” I shut down the phones and restart from a clean slate.
- I swap the sim into the other phone after each Round of testing

Note that whenever I switch the sim card into the other phone and perform the tests, I make sure the phone is put in the exact same location as the previous phone. (Not that it would make much of a difference, but I did it anyway just to keep things constant).

Results of Each Round
During The Morning

Round 1
T-Mobile Server A
Avg Download: 12.744
Avg Upload: 23.776
Avg Ping: 34.8

Verizon Server A
Avg Download: 6.66
Avg Upload:2.784
Avg Ping: 34.8

Round 1 Results:
The Verizon/Qualcomm lost badly on all fronts except for ping, which is tied. T-Mobile wins.

Round 2
T-Mobile Server B
Avg Download: 13.046
Avg Upload: 13.046
Avg Ping: 37.4

Verizon Server B
Avg Download: 9.38
Avg Upload: 1.478
Avg Ping: 37.2

Round 2 Results:
The Verizon loses yet again miserably to T-Mobile and barely wins on ping.

Round 3
T-Mobile Server C
Avg Download: 5.46
Avg Upload: 14.773
Avg Ping: 51

Verizon Server C
Avg Download: 48.33
Avg Upload: 19.45
Avg Ping: 48.33

Round 3 Results:
With the far off server, Verizon takes the lead and obliterates T-Mobile on all fronts!

Round 4
T-Mobile Server A Rematch
Avg Download: 14.71
Avg Upload: 18.62
Avg Ping: 33.33


Verizon Server A Rematch
Avg Download: 14.87
Avg Upload: 27.43
Avg Ping: 37.33

Round 4 Results:
Qualcomm crushes Intel on download speed, and obliterates Intel on all fronts!

Morning Summary:
Out of 4 rounds, the Verizon/Qualcomm model performed better on the download speeds on the latter 2 rounds, the T-Mobile/Intel won the first two. Qualcomm seems to perform better on the "far off" Server C compared to T-Mobile. There is some variability, as in Server A, the Qualcomm lost the first round, and won the rematch.

Afternoon
Round 5
T-Mobile Server A
Avg Download: 14.894
Avg Upload: 23.146
Avg Ping: 36.4
Min-Max Download: 12.11 – 16.48
Min-Max Upload: 21.93-25.69


Verizon Server A
Avg Download: 15.73
Avg Upload: 18.11
Avg Ping: 34.6
Min-Max Download: 11.22 - 17.68
Min-Max Upload: 12.24 - 20.81

Round 5 Results:
Qualcomm had the higher average download speed, though T-Mobile was better on the upload. The max download was higher on the Verizon/Qualcomm.

Round 6
T-Mobile Server B
Avg Download: 11.33
Avg Upload: 19.25
Avg Ping: 38
Max-Min Download: 6.03 - 18.2
Max-Min Upload: 18.64 – 19.61

Verizon Server B
Avg Download: 7.152
Avg Upload: 2.25
Avg Ping: 36.8
Min-Max Download: 5.35 - 9.27
Min-Max Upload: 1.99-2.58

Round 6 Results:
No contest. T-Mobile wins Server B. Verizon wins on average ping.


Round 7
T-Mobile Server B Rematch
Avg Download: 11.692
Avg Upload: 22.718
Avg Ping: 34.2


Verizon Server B Rematch
Avg Ping: 34.6
Avg Download: 15.736
Avg Upload: 17.95

Round 7 Results:
Verizon outperformed the T-Mobile version for download speeds, though T-Mobile beats out Verizon in upload speed. Verizon was barely slower for average ping, but only by 0.2.


Round 8
T-Mobile Server C
Avg Download: 12.608
Avg Upload: 23.614
Avg Ping:44.8

Verizon Server C
Avg Download: 18.208
Avg Upload: 12.104
Avg Ping:47.2

Results of Round 8:
Verizon beat out T-Mobile in download speeds significantly. T-Mobile won on upload speeds by almost double!

Evening
Round 9
T-Mobile Server A
Avg Download: 10.671
Avg Upload: 22.68
Avg Ping: 33.5
Min-Max Download: 5.58 - 14.12
Min-Max Upload: 19.28 - 24.33


Verizon Server A
Avg Download: 15.20
Avg Upload: 17.97
Avg Ping: 32.67
Min-Max Download: 11.74-18.87
Min-Max Upload: 9.43-21.67

Results of Round 9:
Qualcomm wins out again on download and ping, but loses wins on upload speed!


Round 10
T-Mobile Server B
Avg Download: 10.01
Avg Upload: 16.18
Avg Ping: 37.86
Min-Max Download: 4.79-15.00
Min-Max Upload: 1.09-24.06

Verizon Server B
Avg Download: 18.90
Avg Upload: 19.92
Avg Ping: 36.14
Min-Max Download: 14.31-24.02

Round 10 Results:
Verizon wins in all categories, and beat out T-Mobile severely on download this time. It even wins on upload speeds too! The minimums and maximums are higher on Verizon too!

Round 11
T-Mobile Server C
Avg Download: 13.12
Avg Upload: 23.97
Avg Ping: 50.67
Min-Max Download: 10.48-17.91

Verizon Server C
Avg Download: 18.51
Avg Upload: 15.326
Min-Max Download: 14.37 - 21.04

Round 11 Results:
Qualcomm wins out on ping, and crushes Intel on download speed, and beat out Intel severely on download again. Intel makes a comeback with a vengeance on the upload speed.

Congratulations and thank you for reading to the end! Post what you think! Let me know if you have any questions!

 

[ssl0408]

I recently switched to Verizon from AT&T partly because of the Qualcomm difference so it’s cool to see! AT&T has become more expensive than Verizon and their service has gotten worse over the last few years.

 

[Packers1958]

At any time Apple could throttle back the Qualcomm chip to prevent such large discrepancies. Didnt they do that with the 7? My Verizon 7 Plus phone does not get the same speeds as my Verizon X phone.

 

[Starfyre]

At any time Apple could throttle back the Qualcomm chip to prevent such large discrepancies. Didnt they do that with the 7? My Verizon 7 Plus phone does not get the same speeds as my Verizon X phone.

Which one is faster?

 

[KingslayerG5]

Keep it. Best variant of the X there is.

Verizon
Unlocked
Space Gray
Qualcomm modem

Might be the last iPhone to ever carry Qualcomm.

 

[jona2125]

Keep it. Best variant of the X there is.

Verizon
Unlocked
Space Gray
Qualcomm modem

Might be the last iPhone to ever carry Qualcomm.

Wouldn't it be Sprint/Verizon since they are the same model?

 

[rychu]

Keep it. Best variant of the X there is.

Verizon
Unlocked
Space Gray
Qualcomm modem

Might be the last iPhone to ever carry Qualcomm.

Agreed. I returned my AT&T iPhone X and got the Verizon iPhone X!

 

[AmexGG]

Wouldn't it be Sprint/Verizon since they are the same model?

Just having the word Sprint near any phone reduces it's value by at least 10%

 

[Ron21]

NVM, missed the part that said "T-Mobile Sim (same Sim and Service used for each and every test)"

Interesting results and thanks for doing this!

 

[AmexGG]

But.... when comparing the 2 modems, shouldn't you test on the same exact network?

This is more like Verizon vs T-Mobile in your particular area instead of Qualcomm vs Intel.

Please correct me if I'm misunderstanding the data.

Same exact TMO service and sim for every test. Your version of events definitely makes no sense.

 

[lordhamster]

Starfyre,

Thanks for doing these tests, given the resources the average person has to do this testing, I think you did admirable work here.

That said... I’m going to rain on your parade a bit.
I work in the area of Data Analytics and have for over 20 years. I also know how much variability there is in Cellular network traffic, internet routing and speed test sites. At this point, I just don’t feel that there is enough data to draw any conclusions. I feel like there are just too many variables that are too hard to control to make any definitive determination yet.

I often see clients doing some basic analysis, and jumping to the wrong conclusions... in the case of my clients this can cost them a ton of cash. In the case of iPhone buyers, it may lead to unnessesary angst.

 

[Starfyre]

Starfyre,

Thanks for doing these tests, given the resources the average person has to do this testing, I think you did admirable work here.

That said... I’m going to rain on your parade a bit.
I work in the area of Data Analytics and have for over 20 years. I also know how much variability there is in Cellular network traffic, internet routing and speed test sites. At this point, I just don’t feel that there is enough data to draw any conclusions. I feel like there are just too many variables that are too hard to control to make any definitive determination yet.

I often see clients doing some basic analysis, and jumping to the wrong conclusions... in the case of my clients this can cost them a ton of cash. In the case of iPhone buyers, it may lead to unnessesary angst.

I don't work for Qualcomm though I am a techie, I know there is a lot of variability in cellular network traffic. I don't have the $$$ for specialized equipment or the time and means to get a ton of more data to draw hard conclusions, but I wanted hard enough data to draw enough of a semblance of a conclusion for myself. There are too many variables, but I keep what I can constant. I think there is enough data for me to make a decision to lean towards the iPhone I'll be sticking with for myself, and I want everyone else who stumbles on this themselves to help them out too.

I don't want angst from anybody looking at these results that will cause Apple to slow down the Qualcomm modem further, but given posts like this for the iPhone 8:

https://www.macrumors.com/2017/09/28/iphone-8-cellular-bandwidth-tests/

I am certain someone else will do this kind of analysis if I hadn't at some point in time. I can't wait for that. I needed something quicker and reading multiple threads and posts that seem to lean one way or another without any data makes me cringe and want to see for myself what it really was at the time I tested.

I am just a consumer that wants to make as good of a decision as I can with no regrets.
If others have the time to do some tests like this or have access to more expensive equipment, please share results!

 

[gadgetfreaky]

So after reading a lot of articles from reddit posts and articles to forum posts claiming how if your on T-Mobile and AT&T, getting the Verizon model is better because Qualcomm is better, faster, better battery life, etc. I wanted to see for myself how the two compare and determine whether the Verizon or T-Mobile one is better. Hope this helps everyone out with the exact same question and issue as me! Conclusion after the pretty graphs, more about the setup of the experiment and aggregated data/averages below that.

If anyone else has tests and observations they would like to share that you notice from these results, please share!

Pretty Results Up Top (Graphs), 11 Rounds Total, names of each round on left side

Download Summary: Qualcomm beats out T-Mobile/Intel on 8 rounds. Intel wins 3 rounds.
Download Winner: Qualcomm

Ping Summary: Qualcomm beats out T-Mobile/Intel on 7 rounds. Intel wins 3 rounds. Qualcomm is tied with Intel in one round.
Ping Winner: Qualcomm

Upload Speed Summary: Intel wins 8 rounds. Qualcomm beats out T-Mobile/Intel on 3 rounds.
Upload Speed Winner: Intel

Conclusion
I'm leaning towards keeping the Verizon/Qualcomm iPhone X and return the T-Mobile iPhone X because:

1) I think Download Speed and Ping are more important than the Upload Speed.
2) The max and min download speeds (though I don't explicitly include in the post of the Qualcomm when it wins are almost always higher than the min and max of the Intel.
3) T-Mobile dropped cellular data/internet connection in the middle of testing one of the cases and pulling up a web browser didn't work either. Connecting to Wifi, then disabling Wifi did not work either, but after taking the sim out and putting it back in after a minute or two, it started working again. The Verizon did not have such connection issues running through the 11 rounds (each round with 3-6 trials, mostly 5-6 for the majority) I had put it through. Even though it happened once on the Intel, I can confidently say from personal testing, the Qualcomm is more stable, even though it happened on the Intel only one time throughout my tests.

Though I didn't capture any data on battery life, from my observations, after running through the morning rounds, battery life appeared to remain at a 5% decrease on both devices, and measuring some battery life at the end of each round, the battery life appeared to decrease at about the same rate. Others can probably perform some further studies, but I have not been able to conclude that one is more power hungry than the other (at least not to any noticeable degree).

What do you think? Do you think otherwise based on the results? Do high upload speeds matter to you?

Tools Used
iPhone X T-Mobile
iPhone X Verizon
T-Mobile Sim (same Sim and Service used for each and every test)

SpeedTest Server Locations
I used 3 different servers, for the purposes of being anonymous, I am not going to specify which ones they are, but they have the following characteristics:

Server A: Closest to my location in the same state and city
Server B: Slightly close to my location in a neighboring state
Server C: Farthest from my location in another state

Rounds / Times of Test
Each comparison between Verizon/Qualcomm vs T-Mobile/Intel on a single server is a Round. Each Round is an average of 3-6 trial runs at 3 different times in the day (Morning, Afternoon, and Evening). While one phone is being tested, the other is in standby with display off.

Setting Up Experiment

Setup
- Started up both phones, installed Ookla SpeedTest app.
- Auto-Brightness Disabled in iOS 11
- Disable Wifi
- Shut down both phones
- Start up both phones
- Launch Ookla SpeedTest app

Runs
- I manually select Server A, B, or C
- Hit the start/run test button
- Repeat n times for each server
- Whenever I start a new time “Morning/Afternoon/Night” I shut down the phones and restart from a clean slate.
- I swap the sim into the other phone after each Round of testing

Note that whenever I switch the sim card into the other phone and perform the tests, I make sure the phone is put in the exact same location as the previous phone. (Not that it would make much of a difference, but I did it anyway just to keep things constant).

Results of Each Round
During The Morning

Round 1
T-Mobile Server A
Avg Download: 12.744
Avg Upload: 23.776
Avg Ping: 34.8

Verizon Server A
Avg Download: 6.66
Avg Upload:2.784
Avg Ping: 34.8

Round 1 Results:
The Verizon/Qualcomm lost badly on all fronts except for ping, which is tied. T-Mobile wins.

Round 2
T-Mobile Server B
Avg Download: 13.046
Avg Upload: 13.046
Avg Ping: 37.4

Verizon Server B
Avg Download: 9.38
Avg Upload: 1.478
Avg Ping: 37.2

Round 2 Results:
The Verizon loses yet again miserably to T-Mobile and barely wins on ping.

Round 3
T-Mobile Server C
Avg Download: 5.46
Avg Upload: 14.773
Avg Ping: 51

Verizon Server C
Avg Download: 48.33
Avg Upload: 19.45
Avg Ping: 48.33

Round 3 Results:
With the far off server, Verizon takes the lead and obliterates T-Mobile on all fronts!

Round 4
T-Mobile Server A Rematch
Avg Download: 14.71
Avg Upload: 18.62
Avg Ping: 33.33


Verizon Server A Rematch
Avg Download: 14.87
Avg Upload: 27.43
Avg Ping: 37.33

Round 4 Results:
Qualcomm crushes Intel on download speed, and obliterates Intel on all fronts!

Morning Summary:
Out of 4 rounds, the Verizon/Qualcomm model performed better on the download speeds on the latter 2 rounds, the T-Mobile/Intel won the first two. Qualcomm seems to perform better on the "far off" Server C compared to T-Mobile. There is some variability, as in Server A, the Qualcomm lost the first round, and won the rematch.

Afternoon
Round 5
T-Mobile Server A
Avg Download: 14.894
Avg Upload: 23.146
Avg Ping: 36.4
Min-Max Download: 12.11 – 16.48
Min-Max Upload: 21.93-25.69


Verizon Server A
Avg Download: 15.73
Avg Upload: 18.11
Avg Ping: 34.6
Min-Max Download: 11.22 - 17.68
Min-Max Upload: 12.24 - 20.81

Round 5 Results:
Qualcomm had the higher average download speed, though T-Mobile was better on the upload. The max download was higher on the Verizon/Qualcomm.

Round 6
T-Mobile Server B
Avg Download: 11.33
Avg Upload: 19.25
Avg Ping: 38
Max-Min Download: 6.03 - 18.2
Max-Min Upload: 18.64 – 19.61

Verizon Server B
Avg Download: 7.152
Avg Upload: 2.25
Avg Ping: 36.8
Min-Max Download: 5.35 - 9.27
Min-Max Upload: 1.99-2.58

Round 6 Results:
No contest. T-Mobile wins Server B. Verizon wins on average ping.


Round 7
T-Mobile Server B Rematch
Avg Download: 11.692
Avg Upload: 22.718
Avg Ping: 34.2


Verizon Server B Rematch
Avg Ping: 34.6
Avg Download: 15.736
Avg Upload: 17.95

Round 7 Results:
Verizon outperformed the T-Mobile version for download speeds, though T-Mobile beats out Verizon in upload speed. Verizon was barely slower for average ping, but only by 0.2.


Round 8
T-Mobile Server C
Avg Download: 12.608
Avg Upload: 23.614
Avg Ping:44.8

Verizon Server C
Avg Download: 18.208
Avg Upload: 12.104
Avg Ping:47.2

Results of Round 8:
Verizon beat out T-Mobile in download speeds significantly. T-Mobile won on upload speeds by almost double!

Evening
Round 9
T-Mobile Server A
Avg Download: 10.671
Avg Upload: 22.68
Avg Ping: 33.5
Min-Max Download: 5.58 - 14.12
Min-Max Upload: 19.28 - 24.33


Verizon Server A
Avg Download: 15.20
Avg Upload: 17.97
Avg Ping: 32.67
Min-Max Download: 11.74-18.87
Min-Max Upload: 9.43-21.67

Results of Round 9:
Qualcomm wins out again on download and ping, but loses wins on upload speed!


Round 10
T-Mobile Server B
Avg Download: 10.01
Avg Upload: 16.18
Avg Ping: 37.86
Min-Max Download: 4.79-15.00
Min-Max Upload: 1.09-24.06

Verizon Server B
Avg Download: 18.90
Avg Upload: 19.92
Avg Ping: 36.14
Min-Max Download: 14.31-24.02

Round 10 Results:
Verizon wins in all categories, and beat out T-Mobile severely on download this time. It even wins on upload speeds too! The minimums and maximums are higher on Verizon too!

Round 11
T-Mobile Server C
Avg Download: 13.12
Avg Upload: 23.97
Avg Ping: 50.67
Min-Max Download: 10.48-17.91

Verizon Server C
Avg Download: 18.51
Avg Upload: 15.326
Min-Max Download: 14.37 - 21.04

Round 11 Results:
Qualcomm wins out on ping, and crushes Intel on download speed, and beat out Intel severely on download again. Intel makes a comeback with a vengeance on the upload speed.

Congratulations and thank you for reading to the end! Post what you think! Let me know if you have any questions!

- Very cool you did all that. Not a bad test. The DL speed should not be that signficant to be honest. That's a bit surprising.

 

[Gooberton]

At any time Apple could throttle back the Qualcomm chip to prevent such large discrepancies. Didnt they do that with the 7? My Verizon 7 Plus phone does not get the same speeds as my Verizon X phone.

Can they throttle it remotely or do you have to update to say 11.2 or 11.3 or whenever they would do it ?

 

[Packers1958]

Can they throttle it remotely or do you have to update to say 11.2 or 11.3 or whenever they would do it ?

Through a FW update.

 

[Gooberton]

Through a FW update.

So you would have to update manually

 

[Packers1958]

So you would have to update manually

Right. But I'm not saying that's ever going to happen.

 

[walex19]

I initially pre ordered the T-Mobile version but exchanged it at the Apple store for the Verizon model

 

[mhdena]

"
Server A: Closest to my location in the same state and city
Server B: Slightly close to my location in a neighboring state
Server C: Farthest from my location in another state"

What city and state is this?

 

[Starfyre]

"
Server A: Closest to my location in the same state and city
Server B: Slightly close to my location in a neighboring state
Server C: Farthest from my location in another state"

What city and state is this?

I don't want to be specific about where I live, but it is a very populated city and state on the east coast not out in the countryside.

 

[flat five]

I don't want to be specific about where I live, but it is a very populated city and state on the east coast not out in the countryside.

lol..
not really sure if you admitting you live in NYC makes you any more findable to potential interwebz creeperz.

 

[Starfyre]

lol..
not really sure if you admitting you live in NYC makes you any more findable to potential interwebz creeperz.

Your right, but better safe than sorry.

 

[bopajuice]

When I read the title of this forum I thought to myself why did they put their name in the title of the thread? But after reading your comments I have to say thanks for the info, well done.

 

[mjr1000]

Through a FW update.

So is there a way to un-throttle so we can get the gigabit LTE we deserve. (I know our carriers couldn't never reaxh those speeds but you get what I'm asking).

 

[Packers1958]

So is there a way to un-throttle so we can get the gigabit LTE we deserve. (I know our carriers couldn't never reaxh those speeds but you get what I'm asking).

I dont think the chips inside this year’s iPhones support 5G. A FW update would not be possible.