This is our the first smartphone we’ve seen with a 28nm SoC, and thus battery life is the big question. Further, the handset includes all the onboard MSM8960 radio goodness as we’ll mention in a bit. The problem with some HTC phones for the longest time was that they shipped with smaller than average batteries - while the competition continued up past 6 Whr, HTC would ship phones with 5 or so. That changes with the HTC One X, which includes a 6.66 Whr (1800 mAh, 3.7V) internal battery. I’m presenting the same battery capacity chart that we did in the Xolo X900 review for a frame of reference.

Like we did with the X900 review, we’re going to present the normalized battery performance - battery life divided by battery capacity - to give a better idea for how this compares with the competition.

Battery Capacity

As a reminder, the browsing tests happen at 200 nits and consist of a few dozen pages loaded endlessly over WCDMA or WiFi (depending on the test) until the phone powers off. The WiFi hotspot tethering test consists of a single attached client streaming 128 kbps MP3 audio and loading four tabs of the page loading test through the handset over WCDMA with the display off.

Web Browsing (Cellular 3G - EVDO or WCDMA)

HTC is off to an incredible start with our 3G web browsing tests. Even if you assume that Android and iOS are on even footing from a power efficiency standpoint, the HTC One X is easily able to equal Apple's best in terms of battery life. In reality my guess is that the 4S is at a bit of an unfair advantage in this test due to how aggressive iOS/mobile Safari can be about reducing power consumption, but either way the AT&T One X does amazing here. The advantage isn't just because of the larger battery either, if we look at normalized results we see that the One X is simply a more efficient platform than any other Android smartphone we've tested:

Normalized Battery Life - Web Browsing (Cellular 3G)

The Tegra 3 based international One X doesn't do as well. NVIDIA tells us that this is because of differences in software. We'll be testing a newer build of the One X's software to see how much of an improvement there is in the coming days.

Moving onto WiFi battery life the AT&T One X continues to do quite well, although the Droid 4 and RAZR MAXX are both able to deliver longer battery life in this case:

Web Browsing (WiFi)

There are too many variables at play here (panel efficiency, WiFi stack, browser/software stack) to pinpoint why the One X loses its first place position, but it's still an extremely strong performer. Once again we see a noticeable difference between it and the international One X.

Normalized Battery Life - Web Browsing (WiFi)

The big question is how well does the AT&T One X do when we're using the MSM8960's LTE baseband? Pretty darn well, when you consider that it's bested only by the RAZR MAXX with its gargantuan battery. Probably the most notable comparison point here is the HTC Vivid or Galaxy Note on AT&T which both are based on the APQ8060 + MDM9200 combination. 

As a reminder, the Verizon / CDMA2000 LTE devices here are at a bit of a disadvantage due to virtually all of those handsets camping CDMA2000 1x for voice and SMS. The AT&T LTE enabled devices use circuit switched fallback (CSFB) and essentially only camp one air interface at a time, falling back from LTE to WCDMA to exchange a call.

Web Browsing (Cellular 4G WiMAX or LTE)

This particular graph doesn't tell the full story however. In practice the AT&T One X seems to last a lot longer using LTE than any LTE Android phone we've tested in the past. Nipping at the heels of the RAZR MAXX, we need to look at normalized battery life to get an idea of just how efficient the new 28nm LTE enabled SoC is:

Normalized Battery Life - Web Browsing (Cellular 4G)

Now that we have 28nm baseband we've immediately realized some power gains. As time progresses, the rest of the RF chain will also get better. Already most of the LTE power amplifiers vendors have newer generation parts with higher PAE (Power-Added Efficiency), and such improvements will hopefully continue to improve things and gradually bring LTE battery life closer to that of 3G WCDMA or EVDO.

Using the AT&T One X as a WiFi hotspot is also going deliver a pretty great experience:

WiFi Hotspot Battery Life (3G)

WiFi Hotspot Battery Life (4G)

As an LTE hotspot the RAZR MAXX's larger battery is able to deliver a longer run time, however the One X does very well given its battery capacity and size.

Finally, our cellular talk time charts put the AT&T One X in the upper half of our results. Overall the AT&T One X appears to do very well across the board, but it's very strong in the 3G/LTE tests.

Cellular Talk Time

Normalized Battery Life - Cellular Talk Time

 

Software - ICS with Sense 4 Performance
Comments Locked

137 Comments

View All Comments

  • snoozemode - Tuesday, May 1, 2012 - link

    The RGBW in the Atrix is still bad though except for battery performance which is better because of more light coming through.
  • snoozemode - Tuesday, May 1, 2012 - link

    Would be interesting to see battery performance on gps usage, especially on the 8260A and 8960let where the gps is 28nmthat on-die.
  • OCedHrt - Tuesday, May 1, 2012 - link

    Does the best wifi performance not affect battery life? I noticed the One X does have worse battery life on wifi.
  • sprockkets - Tuesday, May 1, 2012 - link

    I don't know but with it checked on an HTC sensation the battery never charged up and that was on the 1amp charger with the phone. Weird. (this was done on the latest Android Revolution firmware for ICS on the sensation.)
  • OCedHrt - Tuesday, May 1, 2012 - link

    Don't they have the same processor w/o the LTE baseband?
  • 3DoubleD - Tuesday, May 1, 2012 - link

    I know this is less important that the other battery life tests you've done, but it would be interesting if you could include some sort of idle power drain test. For one, it would be interesting to see how the Tegra 3 would compare to the S4 here, since the Tegra 3 has the low power companion core for idle tasks.

    Also, you included a color accuracy chart it your iPad3 review, could that be added to the display reviews. It's much easier to interpret than looking at the CIE diagram.

    Thanks for the great review... I'm pretty tempted to buy a white One X after this. I've been impatiently waiting and refreshing Anandtech for weeks now for this review!
  • NeoteriX - Wednesday, May 2, 2012 - link

    The 40nm LP process used for the companion core may be a bit of a gimmick compared to the next generation smaller process fabs. In other words, it's clear that the 40nm LP process cuts down on consumption compared to the 40nm general process (so, compared with the Tegra 2 which is two cores at 40nm general process, there is battery savings).

    However, it could very well be (and is likely) that even the 40nm LP process cannot compete with the 28nm fabrication process used in the Snapdragon S4.

    It would be interesting to see, but my bet would be that the S4 wins even on idle.
  • fm123 - Thursday, May 3, 2012 - link

    The only thing similar is the phone talk time battery life. Since the Krait has an advantage of 3G/4G integration and die size, the talk function largely removes 3G/4G function from the test (which Tegra does not have integrated), and the talk function itself is not too intensive.
  • Stormkroe - Tuesday, May 1, 2012 - link

    I'd like to think that my incessant whining in the mobile forums are responsible for Exynos missing from the glbenchmark charts :) Broken scores, I tell ya!!!
    PLEASE, as a favor to us old 2011 phone owners, re-review phones like the Sensation and Galaxy S2 now that they've gotten ICS.
  • RamarC - Tuesday, May 1, 2012 - link

    battery life vs capacity is useless if the end user can't replace/expand the battery. it doesn't matter if phone A lasts 7 hours with a smaller battery but phone B lasts 9 hours with a bigger battery. if both batteries are fixed, phone B lasts longer, period. now maybe an overall comparison of phone A vs B with all benchies summarized would be good, but even then cost on the same carrier would need to be considered .

Log in

Don't have an account? Sign up now