Galaxy S8 battery life test, Snapdragon vs Exynos edition: a familiar tale

The Samsung Galaxy S8 has the same battery capacity as its predecessor the S7, but lasted longer when we ran it through our standardized battery benchmark. Since both phones come with the power-hungry high-res displays, we can attribute the better endurance mostly to the more frugal Snapdragon or Exynos chipsets of the S8. After all, they are made with the newest 10nm production node, compared to their 14nm siblings in the S7.

While the S8 certainly scores better in that department than the S7, they do share one battery life characteristic - the global version of the S8 with the Exynos chipset still clocks better endurance than the Snapdragon-equipped model for the US and China. It is not a small difference, either - the Galaxy S8 with Snapdragon 835 lasted 7 hours and 30 minutes on our battery test, while the Exynos 8895 model went on almost an hour longer, for 8 hours and 22 minutes. 

While both results mean that the S8 is no "two-day battery" phone like a lot of midrangers with 1080p displays these days, the Exynos would still last you a bit later in the day than the Snapdragon-laden model, with all other things being equal. As to why a 10% or so difference in battery life between the two chipsets, given that they are both 10nm endeavors produced in Samsung's foundries, the answer may lie in the differences in design, as we have custom cores in Samsung's and Qualcomm's silicon, respectively, despite that they are likely still based on ARM's OEM Cortex-A73 and A-53 architecture. 

Another reason may be the fact that the Snapdragon cores are clocked higher than the Exynos ones, all eight of them. While the difference in maximum clock speeds for the faster cores is not that big - Samsung lists 2.3 GHz for the Exynos vs 2.35 GHz for the Snapdragon - the low-power cores are clocked at 1.7 GHz on the S8 with Exynos, against 1.9 GHz on the Snapdragon version for the US, and that might be tipping the scales during our demanding battery test, but may also present a difference in endurance when using the phones as your daily driver.

Related phones

Galaxy S8
  • Display 5.8" 1440 x 2960 pixels
  • Camera 12 MP / 8 MP front
  • Processor Qualcomm Snapdragon 835, Octa-core, 2350 MHz
  • Storage 64 GB + microSDXC
  • Battery 3000 mAh(30h talk time)



1. Tyrion_Lannister unregistered

I just don't trust your battery test anymore. there is no way in hell the S8+ has lower battery life than the S8. Every other site disagrees with you. I highly suggest to re-run the tests. Maybe the S8+ you had was running pre-production software since you got the test done so much earlier.

3. LebronJamesFanboy

Posts: 671; Member since: Mar 23, 2013

*****Daniel P - Serious question for you and other PA authors**** Why do you guys refuse to change your battery test? How have you not noticed that something is extremely wrong with it? You guys refuse to acknowledge that your test is trash, yet you perform it every review. You are literally LYING to potential buyers that use your site for information. You are LYING to the many individuals who turn off their adblock accounts to compensate you for the information you provide. As Tyrion stated, you should know your test sucks if the S8 is somehow beating the S8+. But, the fact you guys won't listen to your loyal readers is shameful.

17. Clars123

Posts: 1079; Member since: Mar 16, 2015

It's not just a matter of doing it for a better review..trash sites like BGR then consequently Yahoo, make articles about Phonearena's battery life tests on the regular.

27. michaelny2001

Posts: 346; Member since: Aug 01, 2012

I don't know much about it, but it coul dbe because the S8+ has a larger screen which eats more battery.

34. cheetah2k

Posts: 2297; Member since: Jan 16, 2011

You're just wasting your time bro. Just ignore this sh*t and go elsewhere for more accurate data.

40. ibend

Posts: 6747; Member since: Sep 30, 2014

meanwhile.. S8+ beat iPhone7plus with ~40% battery left (side note: that i7+ is on airplane mode) in PA's battery test, the only thing that matter is iPhone get 1st place (among all flagships), the rest is just given random number

4. shaineql

Posts: 522; Member since: Apr 28, 2014

Only website on planet where iPhone is competitive in terms of battery. It's a joke, my iPhone 7 goes down up to 50% while watching 2hr movie while s7 edge drops 25%~... It's a joke .

9. Tyrion_Lannister unregistered

Reply to #3(why did the threading get so screwed, it was perfect a couple months back): LCD panels are terrible while viewing videos since the panel self-refresh can't kick in. In browsing though, they are very competitive. iPhone 7 should be within 10% of the S8.

6. AmashAziz

Posts: 2934; Member since: Jun 30, 2014

Why not buy the S8 and create your own script? I think some sites should try a battery life test running this setup: -CSR Racing 2, running at max graphics (in garage screen) -Screen brightness set to 300 nits -Screen timeout: Never -Wifi: On -Airplane/low-power mode: off -All other connections such as bluetooth: off -Night Shift: off -Phone speakers: off -Phone kept on a flat and hard surface, facing front. -Temperature of atmosphere: 35 degree Celsius. Must be an interesting test with these constants, even though a phone might not last more than 3 hours. Would give a true battery life result since gaming is one of the heaviest tasks done on phone. WIth this, tech sites and OEMs won't give fake results of delusional two-day battery life. What if I am rplaying a game continuously, huh?? What would you freaks say then........?

8. Tyrion_Lannister unregistered

I just don't have the money to do such redundant stuff. Let me rephrase it: I have money, but I will lose almost $2000 for something Anandtech will do before I successfully import the SD variant here. Also, i have strong inclination GSMArena and XDA will do the same. Patience is virtue.


Posts: 1168; Member since: Oct 05, 2015

GSMarena's battery test is out for the S8+, and it scored an 88hr rating. The numbers are very similar to the numbers for the exynos Note 7, and not that much higher than the snapdragon Note 7. This tells me two things: a) the battery life gains from the new 10nm doc aren't that big at this stage in the process node's life, and/or b) Samsung still isn't optimizing snapdragon variants very well. The second of those two finding will be confirmed once GSMArena releases an endurance score for the exynos variant.

41. Vador225

Posts: 23; Member since: Jun 19, 2016

But the Screen of the S8+ is bigger (6'2) , with the same 3500mAh of the Note 7 and 100mah less than the S7 edge. And According to GSM arena's Battery Table, it's better than the Mate 9 which packs à 6 inch Display with 4000mAh. It's a really good score I think...


Posts: 1168; Member since: Oct 05, 2015

The screen isn't​ actually much bigger. And with smiled screens, the power draw isn't gonna actually be that much higher with the small amount of extra screen. Also, the S7 family are supposed to have a newer more efficient process and more efficient components, but the numbers are still practically identical to last year's soc. It's not like the gains we got from going SD617 to SD625, those were significant across the board...performance AND battery.


Posts: 1168; Member since: Oct 05, 2015

I meant AMOLED, not smiled. Also, I meant S8, not S7


Posts: 1168; Member since: Oct 05, 2015

That's because Mate 9 has a power hog of a GPU configuration. Put it up against the Mate 8 and its "inefficient" LCD screen and you'll see that last year's microarchitectures and process nodes are still relevant.

28. hitmantb

Posts: 79; Member since: Jul 26, 2012

GSM Arena is pretty consistent as Phone Arena, you just have to remove the phone calls. Their 1 hour phone call / video / web browsing x 3 split doesn't make sense. Just set up a custom score with only web browsing and video. iPhone is always stronger with web browsing because of LCD screen. Galaxy is always stronger with videos because of AMOLED.

29. remixfa

Posts: 14605; Member since: Dec 19, 2008

No crap. Other sites' test show a 2-4 hour battery improvement for the S8+ over the S8. The script they use is crap. Do you guys turn on every samsung option to run these tests, like smart stay and such, which most people turn off and nearly no other phone has? Make sure all options are equal. And run it with something that isnt a WHITE BACKGROUND which is by far the most energy intensive for an AMOLED screen. Hell, do a black background test and see the difference.

35. TheOracle1

Posts: 2340; Member since: May 04, 2015

Why do you even bother? XDA, GSMArena and various other sites are more reliable when it comes to objective assessments. Phonearena is just a tiny step above BGR and the other iSheep sites. Keep in mind that it's a US site so iPhonearena caters to that market whilst receiving their offshore payments from Apple to perpetuate the myth of iphone superiority. I own Xiaomi phones and almost all xiaomi models are in the top 20 of every battery test regardless of what iPhonearena say

2. manojmcn

Posts: 634; Member since: Jul 16, 2015

Galaxy S8 battery life test: A familiar tale - Phone Areas battery test methods are useless.

36. ibend

Posts: 6747; Member since: Sep 30, 2014

meanwhile S8+ beat iPhone7plus with ~40% battery left (side note: that i7+ is on airplane mode, but not for S8+) this site's battery test is pure BS

5. Arch_Fiend

Posts: 3951; Member since: Oct 03, 2015

We all know that Samsung is crap at optimizing snapdragon so of course the exynos version has much better battery life, when the difference probably shouldn't be that big. oh well not as if we can do anything about it.

7. Tyrion_Lannister unregistered

That's not the case. It's just related to the P-states Qualcomm allows Samsung to reach. S7 had the exact same battery life as other SD820 phones on the market. It's just that the E8890 was a much more efficient processor, having the virtue of more efficient low power p-states and 4 A53 cores to dump background tasks on. SD835 has similar issues. The idle frequency(lowest P-state) is much higher than the E8895 and hence it will drain a lot more power when under low load. As you can see, the difference reduced from 60% last year to 13% this year just by using an octa-core configuration instead of quad-core. It's not Samsung's fault, it's qualcomm's fault for not providing competent SoCs. During the S4 days, SD600 model beat out the exynos model in battery life tests because kraits were more efficient than A15s.

12. combatmedic870

Posts: 986; Member since: Sep 02, 2015

So what we will need is a custom kernel to allow for lower frequencies! But alas.... Locked up tight bootloader.

13. Tyrion_Lannister unregistered

I would disagree. There simply isn't a P-state for the processor to downclock to. It's a hardware limit, which is probably related to some engineering issues at Qualcomm. It's really sad, but should still absolutely demolish SD820/821 phones, especially with heavy usage and multitasking.


Posts: 1168; Member since: Oct 05, 2015

I'm not sure the difference will be that great. Without trying too hard, I can get 9 hours sot with the 3T on the SD821 (2.34ghz/2.18ghz). I bet if I stopped the max clocks to something like 2.15ghz/1.28ghz it would fair just as well in real world but give even better battery.

19. Arch_Fiend

Posts: 3951; Member since: Oct 03, 2015

It's not the whole case true, but the fact still remains that Samsung does a poor job at optimizing snapdragon chips sets when compared to LG and HTC handsets with the same chip, like last year with the SD820. Sure the Exynos SOCs of today are usually more efficient than Snapdragon but not as much as the difference in battery life between Samsung's own handsets would suggest.

21. tedkord

Posts: 17464; Member since: Jun 17, 2009

That then makes it Samsung's fault for using QC SOCs when they could use the Exynos.

30. remixfa

Posts: 14605; Member since: Dec 19, 2008

They can't in the USA. U think they wouldn't use their chip and over a snapdragon if they weren't forced? Qualcomm's patent trolling is legendary.

48. vincelongman

Posts: 5750; Member since: Feb 10, 2013

Samsung can use their Exynos if they really want They just have to pay the licensing like Apple and what they did with the S6 But they aren't since its cheaper to use the 835 in the US

10. LebronJamesFanboy

Posts: 671; Member since: Mar 23, 2013

PhoneArena 2017 Review + Awesome Community + Recent content; interesting articles - Battery tests are awful - Commenting has been broken for too long - TechieXP (jk)

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit for samples and additional information.