Figured I'd share my data I gathered during the past couple of days or so. Zip has all the screenshots.
I ran PCMark's Work 2.0 and 3DMark's Sling Shot Extreme Vulkan benchmarks back to back on my Xperia XZ1 Compact to see how well Sony's thermal management is working on the XZ1C.
Spoiler alert: really well.
PCMark Work 2.0: nothing to see here. SD835 is quite efficient so there's practically no performance degradation over time. Battery temperature starting point was 6*C and it stabilized around 29-31C which it reached during round 6 or about 45 minutes in to the test. Didn't see any point in going any longer than 15 rounds as the battery temperature didn't continue rising.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
3DMark Sling Shot Extreme using the Vulkan API: here we actually see something. Starting battery temperature was 5*C this time.
The XZ1C maintained top performance for 3-4 rounds (12-16 minutes) before starting to rather gracefully slow things down. It reached steady state on round 10 or 40 minutes in to the test. After that we see the beginning of "the great oscillation".
Performance jumps back up for tests 17-19 as the phone cooled down slightly during tests 10-16 before dropping back down to steady state on test 20.
I quite like how Futuremark is giving us the physics FPS numbers in three parts, if you take a closer look you can see really well that the phone is doing in regards to CPU clockspeeds.
And here's another graph for the 3DMark test, this time showing the battery temperature over time as well as the percentage of performance retained from the first test. Here we can see that in a worst case scenario we retain 90% of graphics performance while the CPU nosedives to ~60%.
Related
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
As the graph shows... Not connected to the charger yet it increases 5% in battery life, went from 29 to 34...
Sent from my Galaxy Nexus using XDA App
Same
Mine does the same actually. Everytime I look at it when its below 50% life it always has a part of the graph as it charged a little bit. seems wierd
The battery usage and percentage in the notification panel is more of a guess of where your battery really is based on usage. A lot of usage will cause it to guess your battery is at a certain level, and then less usage will lead it to believe your battery is at a higher level.
It also has to do with the charge going to the battery, but that's a bit more involved.
Powell730 said:
As the graph shows... Not connected to the charger yet it increases 5% in battery life, went from 29 to 34...
Sent from my Galaxy Nexus using XDA App
Click to expand...
Click to collapse
This is normal behavior for a battery, i've had this with all my android phones (HTC Magic/Galaxy S/Galaxy Nexus).
I have that too. But I believe Tesla hasn't died and is currently working with Google or Samsung on some sort.
Poorly calibrated battery or a big change in ambient temperatures?
For example here in Finland, it can be like -20'C outside during wintertime and the battery seems to take a hit when using the phone outside and after coming inside to +20'C (or more) it could suddenly (when warming up) load the battery magically
Mine does the same and I love it!
this is normal for this type of battery fuel gauge because it's based soley on voltage. and you pressed the phone hard during that steep slope, then went idle, and voltage tends to rise and bounce back up when in a relaxed state. when this bounce is large enough it moves the SoC as well. it's all normal.
So warm up check is something you can do the moment you get the phone (or at least it was like this with the Sensation).
Can someone that has already received the phone comment if the phone has a warm up problem while playing games or web-browsing etc?
(I know most of the phones warm up, the question is if it's fine or if it's becoming an oven like the early versions of the Sensation)
Been toying with it for about an hour now, battery just about to die so I can charge properly. No heating issues what so ever. It is as cool as ice tbh. I've played Glowball a bit and the screen has obviously been on for an hour now, 6% battery left, started with 27. Wifi on, data on, heavy game on. It does NOT heat up and Glowball runs perfectly smooth for me. I will be doing more tests in a day or 2.
Sent from my HTC One X using XDA
djolebih said:
Been toying with it for about an hour now, battery just about to die so I can charge properly. No heating issues what so ever. It is as cool as ice tbh. I've played Glowball a bit and the screen has obviously been on for an hour now, 6% battery left, started with 27. Wifi on, data on, heavy game on. It does NOT heat up and Glowball runs perfectly smooth for me. I will be doing more tests in a day or 2.
Sent from my HTC One X using XDA
Click to expand...
Click to collapse
thanks for confirming glowball
can't wait for your full hands on
Thank you! Waiting for more comments
Phone:=HTC One X
Frontpage:=28,1°C
Shadowgun:=37,4°C
Shadowgun + charger:=40,9°C
Got this from a dutch review
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
I tried the latest antutu bench mark test. I usually got 5-6k score but now I had this score
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
blackwatch92 said:
I tried the latest antutu bench mark test. I usually got 5-6k score but now I had this score
Click to expand...
Click to collapse
I bet u got huge increased score on 3D, this happened to me also. It is bug on Antutu i guess. Ur 3d test didnt really went smooth,right? i mean the screen just stutter for some sec and test test complete LOL. just ignore it, NO INCREASE IN PERFORMANCE OR APP HACK ITS BUG I BELIEVE
And i normally get around 6k only!!
Thx that explains it.
lol
i have just got 4000 XD
Lol ThreadBump:laugh: Yeah Benchmarks are only a bit trustable like to see speed or whut but whats more important is the Smoothness you experience, Benchmark is just some computer logging, Nothing is more important then the feeling you have about it since nowadays Manafactures are cheating on benchmarks with benchmark detection software and then having everything on Max power for the whole time.
Btw I didn't made any new progress yet on the Fame Cm10/Cm11 project, I am busy with School and my personal life. I am trying to continue it when i have time but you guys are free to continue also the project and use my sources since I don't have the time of the world to finish it at once :/ So I will still try to work on it :good:
Howdy...
I've been wondering how the Samsung S9 with Snapdragon behaves when throttling.
it's very easy to test... just run this app called CPU throttling Test and take a look at Temperature, Clock and Performance.
This is a Screenshot from my Sony Xperia XZ2:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
You can clearly see the Temperature reaching a critical point and the Performance dropping, while Frequency stays at the same 2.8ghz.
Here we have a OP 6 which does show the same behaviour... no Frequency Limit!
Looking at the SD835 Essential Phone:
or the even older SD820 HTC 10:
We do see a CPU Frequency Limit on both devices.
So... is this Max Frequency yet still reduced Performance a Snapdragon 845 thingy?
Simply do the Test for 2-5 minutes (until throttling becomes obvious) and we will know.
P.S going to Options and chaging Threads from 20 to 7 might totally change the behaviour of this test!
P.P.S Go into Options and correctly set Thermal Zones! If you don't know which nr is what... Terminal / ADB shell "for file in sys/class/thermal/thermal_zone*/type; do echo "$file"; cat "$file"; done"
Greetings,
Haldi
Somebody made a video with 3 SD 845 phones, but he got higher performance, did you change number of threads from default 20 and where phones cold when you begin testing?
I can't post a link, just search "OnePlus 6 vs Sony XZ2 vs Xiaomi Black Shark" on youtube.
Hello, in the recent days I've come to notice that my p6p takes more time to charge (when charging with the same cable(factory cable), charger(pd3.0 pps) from the same battery percentage), it now need almost 3 hours to fully charge from 15 to 100% when it used to do it in 2ish. The data in the apps accubattery and battery guru confirm this, since I can compare the charge times in different charging sessions.
When inspecting the wattage and current draw while charging ( 15%-100%) I can see a clear change in max sustained wattage, from the usual with p6p 22ish watts now it only charges at 17ish. Th current is also decreased from 4.5 to 3.7 Amps (at 5v input read, I know that software readings can be off, like in this case, since pps doesnt do constant 5v 4.5A) still while using the same setup at the same temperature the behavior is different, Such difference can be observed in the current vs time graph from the battery guru app. More over the current doesn't stay constant, it fluctuates between 0 and some charging value all the time( the app constantly changes between charging and not charging) this can been seen in the graph, as the "now" graph is more pink that indicates the jitter in the current while the "before" one in more transparent and the current is constant.
Typical curve some months ago
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Typical curve now
Can this be a software issue or should I go and buy a new charger/ cable?
Has anyone else have encountered this behavior ?