Last week, we presented you with a quick preview of the performance of the Samsung Galaxy S20 Ultra based on Snapdragon 865, showing that the phone has some extraordinarily good performance and energy efficiency statistics. Since then, we’ve managed to get our hands on an Exynos 990-based Galaxy S20 + and S20 Ultra, subjecting the trio of phones to our usual extensive overhaul process.
We are still working on the great full article which is still a week or more away, but we already wanted to cover one of the main talking points about the new devices: battery life. Samsung’s new 120Hz update mode is quite a nasty hunger, which alters the battery life formula for this year’s flagships. In addition, we are still experiencing some fairly large differences between Exynos and Snapdragon based phones and we are able to report the first preliminary battery test results and analyze the appearance of the situation.
Editor’s Note: The current results have been tested on the ATBM firmware for the Exynos variants and ATBN for the Snapdragon variant.
Very expensive 120Hz mode, power degradation across the board
A key aspect of the new Galaxy S20 devices is their new 120Hz screen display mode. Samsung’s new high refresh rate is undoubtedly a killer feature for new flagships, and when it comes to the performance and fluidity it offers, it’s an incomparable experience that only very few devices can compete with.
To achieve this performance, Samsung seems to have had to make some sacrifices regarding the energy consumption of the phones. While Samsung managed to implement 120Hz mode without any additional hardware such as additional MIPI interfaces between the SoC and the display controller IC (therefore, because it is limited to 1080p 120Hz), effective power management and absorption of energy when in 120Hz mode it is extremely inefficient and poorly optimized.
On both Snapdragon and Exynos units there is evidence in the display drivers that there is hardware support for some sort of VRR (variable refresh rate) – what exactly this VRR implementation represents at the moment I don’t have any more information, but it doesn’t look like it is a “real” VRR mode, at least not in the current state of affairs of the software. Samsung has implemented the implementation of the panel with apparently 4 different update frequencies; 48Hz, 60Hz, 96Hz and 120Hz. Apparently, is it possible that this VRR implementation is switching from one mode to another?
There is also evidence of an unfinished software stack; apparently Samsung had planned some sort of adaptive mode at 120Hz, however this currently doesn’t seem to be active and the phone simply stays at 120Hz all the time.
Galaxy S20 Ultra (Exynos 990) – Device power 120Hz vs 60Hz (black screen)
The worrying thing here is that 120Hz mode has a rather large power impact that is always on, regardless of the screen content. Above, for example, we see the basic power of the S20 Ultra with Exynos 990. Both areas highlighted are on a static black screen and there is a huge difference of ~ 184mW between the two modes. 184mW may not seem like much, but since it is a basic consumption of the phone that is always present every time the screen is turned on, it adds a considerable amount of energy on an entire battery charge.
This delta between 60 and 120Hz mode is present on all S20 devices that I have here at hand, both on Exynos 990 S20 + and S20 Ultra, and on Snapdragon 865 S20 Ultra:
In addition to the huge impact of 120Hz mode on the basic power consumption of phones, we also see that there are greater power degradations even in 60Hz mode. Exynos 990 here is around 40-100mW worse than the Exynos 9820 powered S10 +, depending on the measurement on S20 + or S20 Ultra.
What is actually more worrying and initially made me very pessimistic about the S20 series is however that the Snapdragon 865 on the S20 Ultra has even more horrible basic power consumption figures, showing ~ 150mW degradation compared to the Snapdragon 855 S10 +.
It is interesting enough to see the Exynos 990 S20 Ultra being far ahead of the Snapdragon 865 S20 Ultra in this metric. I’m not sure exactly what the cause is here, if that variant of the phone is more optimized or simply a hardware feature of the SoCs (the Snapdragon 865 has a 4MB L3 CPU always on compared to 1MB on Exynos 990).
I also included the ASUS ROG Phone II in the table, which shows what 120Hz impact it had on that phone, with equally large degradations on a pure static black screen.
Battery life resulting
We will cover the hardware aspects of the device trio more fully in the full review, but for now I wanted to present our preliminary battery life data. For the results, we expect some confusion of results regarding the various devices and the display modes.
The Galaxy S20 Ultra comes with a 5000mAh battery which represents a 22% increase compared to the 4100mAh unit of the Galaxy S10 +, but it is also equipped with a larger screen and we have seen that the basic energy consumption has decreased a lot , especially in 120Hz mode. The 4500 mAh battery of the Galaxy S20 + is 9.7% larger than the S10 +, but it also has a higher screen.
Above all, we have seen that the Snapdragon 865 has excellent computing power efficiency which should help that variant of the S20 Ultra in our test here. As for the Exynos 990 variants of the S10 + and S20 Ultra, we will cover this topic much more widely in the full review, but apparently SLSI has failed to improve the energy efficiency of its M5 cores, resulting in an energy efficiency deficit of 100 % compared to the A77 core of the Snapdragon 865. The A76 cores of the chip, while ahead of the M5 cores in efficiency, are also far behind the central A77 cores of the Snapdragon 865.
Finally, the following battery life data were performed in the 60Hz and 120Hz FHD modes, using the device’s “High Performance” operating mode. In the latter case, I noticed that for the S20 this year this mode seems to have a greater impact on energy consumption, particularly on the Exynos 990 variants of the phones, something that was not so widespread in the S10 series. We will repeat the battery tests in “Optimized” mode for the complete review, in case there are substantial differences.
In the web test, we start with the Snapdragon 865 Galaxy S20 Ultra. In 60Hz mode, this phone manages to achieve the maximum longevity that we have seen from a flagship of Samsung and the energy efficiency of the SoC and the 5000 mAh battery is able to push the device in front of the S855 S10 + over an hour, even in the face of the degradation of the basic energy consumption of the phone, which is apparently more than offset by the high energy efficiency of the SoC.
Switching to 120Hz mode results in a 20% battery life penalty. The degradation here is among the worst in the high refresh rate display group, similar to what the Asus ROG Phone II showed when switching between its 60 and 120Hz modes.
The Exynos 990 Galaxy S20 Ultra falls behind its brother Snapdragon. It is 13% and 11% lower respectively in 60Hz and 120Hz results. It’s not too dramatic a difference, but it’s enough to put this variant of the S20 Ultra behind the S10 +’s battery life data.
The S20 + with Exynos 990 further reduces the battery life of the S20 Ultra by 10-13%. This is where we see the generational differences between S10 + and S20 + for this phone chipset variant being quite disappointing. Even with a 10% larger battery, the S20 + shows 15% worse battery life than the Exynos 9820-based S10 +. 120Hz mode has a similar 20% battery life penalty that we’ve seen on other phones. In terms of absolute battery life, the S20 + in 120Hz mode comes close to figures that we would consider rather negative for a flagship phone in 2020, resulting in figures under 9 hours in our test.
PCMark’s battery life test is heavier than our web test, mixing some heavier CPU and GPU workloads.
The Snapdragon 865 S20 Ultra worked excellently in both 60 and 120Hz modes – the results end up where we expected them, however the 25% degradation of the 120Hz mode surprised us a little. It is an excellent overall result here in terms of absolute numbers.
The Exynos S20 + in 60Hz mode managed to stay in line with the S10 + last year in minutes. It’s not a terrible result, but I guess it would be a little worse than a Snapdragon S20 + (we don’t have that variant).
In 120Hz modes, both Exynos variants of the S20 + and S20 Ultra show some pretty terrible battery life data. For the S20U, the Exynos variant here falls within the Snapdragon unit by a whopping 27%, probably due to the vast energy efficiency deficit of the Exynos SoC. In absolute numbers, these are among the worst data we have tested in recent years, approaching the disaster that has been the Exynos 9810 S9 series.
I had only received the E990 S20 Ultra last Friday and since then I have run back-to-back battery tests since then, at the moment I am missing the 60Hz PCMark figure for that phone, which I will add to the table as soon as I have completed that test . I must also note that, since I noticed a much worse power hit in “High Performance” on the E990 variants of the phones compared to the S865, I will run the figures again in the “Optimized” mode to check for any important variations (also in this case our figures S9 and S10 were in performance mode).
Paint an image of two different phones
What our preliminary battery life data show today are two important takeaway aspects for the S20 series:
Samsung’s 120Hz mode is a very energy hungry mode that apparently isn’t optimized as hoped for. While the resulting performance and fluidity are exceptional, expect an average hit of 20-25% on battery life depending on the use case. People who use their phones with higher brightness levels will see a lower percentage degradation, while those who prefer lower brightness levels will see greater degradation, due to the higher basic energy consumption of the phones.
Second, the S20’s Exynos 990 variants aren’t doing very well compared to the Snapdragon 865 variants. While last year Samsung had managed to show the rough equivalence between the Exynos 9820 and Snapdragon 855 variants when it came to battery life, this year, the gap in battery life between the two is again noticeable. We will re-run some of the battery tests presented here and hopefully they will improve a little in their results, but the only thing that is clear is that Samsung’s SoC and CPU cores are behind some very significant margins.
We will dwell in more detail on a post mortem of Samsung’s custom M5 cores in the full review, as well as covering any differences in camera quality between the two phone chipset variants, along with all the other usual aspects of the full phone review .
For people whose purchasing decisions are heavily influenced by the battery life of the devices, I think it’s safe to say that the Exynos S20 variants will likely end up being disappointing. Experienced users, especially those who use heavy apps, will see very disappointing results: I would probably recommend importing a Snapdragon unit or simply skipping this generation altogether.