T-Mobile smartphones have longer battery life, Laptop Mag test shows

August 7, 2014

Galaxy Note 3

We often talk about battery size, screen resolution, and SoC as factors that determine how much battery life we can get out of our smartphones. But there are other factors at play, and some of them, like the network you’re using your phone on, are quite subtle.

Case in point: Laptop Magazine battery life tests show that devices running on T-Mobile’s network consistently last longer than the same devices on competing networks, sometimes by a big margin. The pattern is visible “again and again across multiple handsets” according to Laptop Mag’s Michael Prospero.

battery life test

To come to these numbers, Laptop Magazine ran a benchmark consisting of loading 50 popular websites and pausing for one minute on each page, until the battery is drained. The test was repeated on each device-carrier combination, with Wi-Fi, Bluetooth, and NFC turned off and the screen set to the same brightness of 150 nits.

In order to account for the particularities of a specific test location, testers ran the benchmark in New York and Chicago, while ensuring that the phones had at least three signal bars at all times.

T-Mobile came on top in all tests, and in some cases, like the One (M7), it beat the last place by a full three hours.

Why this discrepancy? First, we can’t rule out a flaw in the benchmark – it’s possible that, by coincidence, T-Mobile has better/less congested coverage in the two test locations. A comprehensive test would measure battery life across dozens of locations, both in cities and in rural areas.

Assuming that the test was not flawed, it may be that T-Mobile’s LTE network is less demanding on devices or that T-Mobile loads less bloatware or even device monitoring services on its devices.

Have you ever noticed a difference in battery life across networks?

Comments

  • Acea Alex

    well verizon is worst carrier ever…too much bloatware…software and hardware. Its just crap all around. Well I’m from europe so yeah here there isn;t such thing as carrier hardware bloatware…nor software

  • gg

    There are hundreds of network-related factors, features and parameters that can affect the battery life of a smartphone, and these tests were not exhaustive enough (by a long shot) to properly assess any of them.

    “We also turn off Wi-Fi, so that the phone uses only the carrier’s network, and make sure that it’s receiving at least 3 bars of service.

    3 bars of what service? CDMA? GSM? EDGE? UMTS? HSPA? LTE? Each of those will put a different battery draw on your phone.”

    Also, “3 bars”? Seriously?

    I understand you’re not network engineers and probably don’t have the proper software, but couldn’t you at least look at this screen[1] and report the actual signal level instead of a completely innacurate 4-step estimation?

    For example, signal level on LTE is usually between around -140 dBm (very poor) and around -40 dBm (excellent). You could be inside the same “bar” and still have very different signal levels, and thus very different power levels on the phone’s amplifier.

    “After compiling these results, we wondered whether the delta in battery life was limited to the New York / New Jersey area where we had run all of our prior tests. So we sent all four carrier versions of the Galaxy S5 to a tester in Chicago who also ran our test, with similar results.”

    Again, this shows very poor judgement.

    2 single data points means nothing when comparing 4 different mobile networks. Point 1 may be good for operator A and B and bad for operator C and D, while Point 2 could be a completely different thing. The fact that they tested 2 points and the results were similar means very little without actually measuring other parameters.

    The correct way to do this would be to look for a point with a fixed signal level on operator A (i.e.: -90 dBm), do the test, then look for another point with the same signal level for operator B, then the same for operator C, etc.

    Furthermore, as I said there are hundreds of parameters on the network (not on the device) that will affect battery life, besides coverage.

    For example, inactivity timers might be different in each network (and in each area) for lots of reasons… which means that your phone stays fully connected to the network for a longer time before going to sleep.

    Also, networks will tell the device to start scanning nearby cells when your current signal level drops below X dBm, in order to prepare for handover when the current cell is no longer suitable. That X level might be higher or lower depending on the operator, the frequency band, the area, the technology, etc. Every single criteria is configurable, and neither is better or worse than the other – there are many different strategies.

    A lower value means that you’ll stay longer on your current cell before handing over, but also that you’ll be on worse coverage for longer. Maybe it’s worth staying longer on 4G because your alternative is a piece of shit called CDMA, so no matter how bad the signal is, LTE will be better. Or maybe you have HSPA+ and you really prefer to handover sooner, as a low LTE signal will be worse than a good HSPA one. Scanning will put a lot of load on the battery.

    There are also lots of power control features that actually tell the device how to behave with regards to the power amplifier and signal strength. Yes, the network can actually tell the device how “loud” it should (or shouldn’t) be depending on many conditions.

    Finally, different networks have different infrastructure vendors. Your network might be called “Verizon”, “T-Mobile” or whatever, but in reality your network is manufactured by Huawei[2] , Ericsson[3] , Alcatel-Lucent[4] , etc., (both hardware and software). The vendor could even change inside the same operator, depending on the area and of course the technology (2G, 3G, etc.)

    Different vendors have a different roadmaps, different features and different performance, which can affect the interaction with the phone and the impact on battery life.

    For example, one vendor might support DRX[5] and the operator may use it, allowing for massive battery savings. While other vendor might not support it yet or your operator could have decided not to enable it for all kinds of reasons. There are lots and lots of features that can be on/off, or parameters that might be configured with a different value, etc.

    No matter how “consistent” they say this trend was, they probably did most of the tests in the same area and without any proper measurements… so it’s very, very difficult to draw any conclusions from this.

    • Vomitous Mass

      “possibly”
      “maybe”
      “could have”
      “depending”
      “probably”

      Wow – if you used any more CYA words, you’d get elected!

      To those new to the internet:
      Use of these words protects the author from being called out as “incorrect” while allowing the author to imply his opponent is incorrect or false without having to actually make the claim or back it up with actual fact/evidence. Any post containing a large quantity of them should be seen as it is: lots of words – no content.

    • TMOTECH

      -140dBm? Are you sure about that number? Try again.

    • reader

      Yeah, you seem so desperate in your explations I didn’t bother to read past the first paragraph of your long essay. All I know is you probably work at the one the listed carriers other than T-Mobile.

  • MasterMuffin

    Must be a flawed test. If the #1 was Verizon, I’d say the testers were paid (:D), but because it’s T-Mobile, I think there was just some factor(s) they didn’t notice

  • derp hurr-durr

    The sheer desperation of the attempts to “explain” this away as a non-benefit is incredibly amusing. They’re crawling out of the woodwork to find any excuse, no matter how ridiculous or untrue, to deny T-Mobile any credit for this.

    This article in Laptop Mag really touched a nerve. Didn’t realize T-Mo had quite that many haters.

    • Caleb Wright

      There is always going to be haters for every carrier that’s never a surprise. The only difference is that what Bill suits your financial needs some people just can’t afford Verizon or AT&T, and then some people don’t like the dropped calls or 2g network that some people experience on tmobile. And Sprint I’m not getting into their hidden features and backlashing crap.

  • Caleb Wright

    We already discussed this on phone arena it’s not very conclusive results because tmobile connects to 2g networks where verizon has 3g and 4g only so 4g will kill the battery faster then 2g for sure so you can’t dis Verizon for that reason and if you are smart enough to root your phone that takes off the bloatware. So those two things are irrelevant for battery life you turn off mobile data and take away bloat ware then the battery should last exactly the same.

  • HeyRadar

    They should have included iPhones in the study.

  • Mdraksi

    Its not in batery. It is in network refreshing rate,quality of signals and coverage and most important thing- equipement. TMobile always buy new stuff,and not some used Alcatel centrals from XIX century

  • Harry

    it would have been interesting to see the battery profile on each run. Another would have been to verify the capacity before the test (although the latter would have negligible impact if manufacturers maintain tight quality controls, I would like to think). Also it is a function of the frequencies used where the test was run.

  • smokebomb

    I can’t fairly compare between Verizon and T-Mobile, but I know for a fact I needed almost 2 batteries to get through a regular day when I had the galaxy nexus on Verizon. I even bought a gigantic battery because I was tired of swapping. That one lasted a day with moderate use. That gigantic battery is the same size as the stock one that comes with the note 3.

    I have a note 3 on T-Mobile. It lasts from the minute I wake up to the time I go to bed (6 am – midnight) unless it’s a day where I ride the bus. On those days, I kill 20% of the battery watching videos and reading online in an hour.