Antutu Benchmarking Problem - Xiaomi Mi Mix 2S Questions & Answers

I have just bought this Mi MIX 2S yesterday, ad after an update I tried to benchmark the device using Antutu. Somehow, after numerous trials, I could only get the score hovering between 155.000 - 175.000 (versus the anticipated 260.000). I even tried turning off all connection and rebooting. Do anyone encountered the same issue and is there any justifiable cause for this low benchmark score ?
Thanks in advance.
*Screenshots attached for reference*
{
"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 have same problem,score 180000 ,maybe bug of last version on antutu?

Just don't care about the antutu score... Do you notice a difference in daily use? If not than it's not important that antutu shows a different score

Flatric said:
Just don't care about the antutu score... Do you notice a difference in daily use? If not than it's not important that antutu shows a different score
Click to expand...
Click to collapse
Agree, benchmark are only for fun, real use is more important. Consider that benchmark can be influenced from temperature.

arkemens said:
i have same problem,score 180000 ,maybe bug of last version on antutu?
Click to expand...
Click to collapse
Might be the firmware hindering the app from dragging highest device performance

...Put your phone for 5 minutes in the fridge. Start the test. Keep it in the fridge while the test runs. Take it out of the fridge. And voilà: a new high score star is born...

Getting low score on 3dmark (3K instead of 4500) and when I play highest graphics on PUBG Mobile it's not smooth as other Snapdragon 845 devices

feicher said:
...Put your phone for 5 minutes in the fridge. Start the test. Keep it in the fridge while the test runs. Take it out of the fridge. And voilà: a new high score star is born...
Click to expand...
Click to collapse
Yes that the reason why i mentioned temperature, i tried it on my old device but now not needed and received over 257000 points on Antutu.

VampireHeart said:
Yes that the reason why i mentioned temperature, i tried it on my old device but now not needed and received over 257000 points on Antutu.
Click to expand...
Click to collapse
I have the same problem with rom 9.5.20.0
It is a real performance problem, it affects CPU but not GPU
Just receiving it had the correct performance, after a few days under drastically and to test it against Geekbench, it gave me less than 2000 single core and less than 7000 multi core (normal values ??for that test in mix 2s are 2400+ and 9000+ )
I did a hard reset and received the normal points and adequate performance for a few days (even starting at 37 degrees, the temperature has nothing to do in this matter)
Notice slight drop in performance and pass the tests again and again it did not exceed 180k antutu and 2000/7000 GeekBench, the mobile started completely cold
It gives me less CPU performance than a mobile with Snapdragon 820 that I have out there and I feel scammed.
If this is not solved with MIUI 10 it will be the last Xiaomi device that I bought, I already sold an MI5S because with updates I turned off 2 cores with a super restrictive thermal and even so it got hotter and definitely worked worse
I'm very disappointed

macho3 said:
I have the same problem with rom 9.5.20.0
It is a real performance problem, it affects CPU but not GPU
Just receiving it had the correct performance, after a few days under drastically and to test it against Geekbench, it gave me less than 2000 single core and less than 7000 multi core (normal values ??for that test in mix 2s are 2400+ and 9000+ )
I did a hard reset and received the normal points and adequate performance for a few days (even starting at 37 degrees, the temperature has nothing to do in this matter)
Notice slight drop in performance and pass the tests again and again it did not exceed 180k antutu and 2000/7000 GeekBench, the mobile started completely cold
It gives me less CPU performance than a mobile with Snapdragon 820 that I have out there and I feel scammed.
If this is not solved with MIUI 10 it will be the last Xiaomi device that I bought, I already sold an MI5S because with updates I turned off 2 cores with a super restrictive thermal and even so it got hotter and definitely worked worse
I'm very disappointed
Click to expand...
Click to collapse
Like said previously benchmark are only for joke, you must see effective performance in daily use, I tried now another test with latest Xiaomi EU build and received around 240.000 points, yesterday same build around 194.000, now over 267.000.
I had that result without modify nothing so I think depend from many factors, like temperature, apps running in background and other.

Incredible...."Antutu Benchmarking Problem". WOW!
Yesterday I saw someone posting something like "broken performance". Geeeez.
People ignore what they see in person and rather believe in a program KNOWN to be manipuluted as hell by manufacturers.
Back then they all said Antutu is Bull****. It seems since the day they learned how "to play" Antutu...they are even going public with numbers they created.
No doubt....it´s always good and nice to see the performance you feel in numbers by a benchmark. But please people stop the madness and THINK about it.
Other than that 9.5.19+20 are known to give bad results with antutu.
I couldn´t care less. Best proof for antutu being nonsense is...install MIUI 10 latest Beta....and feel the speed of it....(it is just a little tad snappier to me). But Antutu Score is pretty stable 260tsd+ !
(mine is 128gb/6gb)
Anyway I´d advice you to try that Beta....I find it really decent and has really nice new features.
You may happend to have a different use case and experience bad bad problems not yet fixed...then bad luck for you. I´m really lucky with it currently.

DJ-Tumor said:
Incredible...."Antutu Benchmarking Problem". WOW!
Yesterday I saw someone posting something like "broken performance". Geeeez.
People ignore what they see in person and rather believe in a program KNOWN to be manipuluted as hell by manufacturers.
Back then they all said Antutu is Bull****. It seems since the day they learned how "to play" Antutu...they are even going public with numbers they created.
No doubt....it´s always good and nice to see the performance you feel in numbers by a benchmark. But please people stop the madness and THINK about it.
Other than that 9.5.19+20 are known to give bad results with antutu.
I couldn´t care less. Best proof for antutu being nonsense is...install MIUI 10 latest Beta....and feel the speed of it....(it is just a little tad snappier to me). But Antutu Score is pretty stable 260tsd+ !
(mine is 128gb/6gb)
Anyway I´d advice you to try that Beta....I find it really decent and has really nice new features.
You may happend to have a different use case and experience bad bad problems not yet fixed...then bad luck for you. I´m really lucky with it currently.
Click to expand...
Click to collapse
Agree for all the words, benchmarks are only a joke, i tried also on other devices like OnePlus 5, with some roms and kernels i had less points but the device is fast like hell.

Thanks guys for all the help and opinions that you expressed ^^
After a while I stopped giving crap about benchmarking (except when I want to fiddle with ROMs of course). I opted in for Android P beta and I got like 280k+ so really I'm basically relieved and going all-in testing the device
I also agreed that MIUI 10 is pure satisfaction too ~~~ can't wait for the build that based on Android P :3

I know how to fix
Everyone should read my thread: (Rooted) improve performance. I tried and successed!
---------- Post added at 05:51 PM ---------- Previous post was at 05:43 PM ----------
TheSacredKey said:
I have just bought this Mi MIX 2S yesterday, ad after an update I tried to benchmark the device using Antutu. Somehow, after numerous trials, I could only get the score hovering between 155.000 - 175.000 (versus the anticipated 260.000). I even tried turning off all connection and rebooting. Do anyone encountered the same issue and is there any justifiable cause for this low benchmark score ?
Thanks in advance.
*Screenshots attached for reference*
Click to expand...
Click to collapse
I fixed. But my phone may be hotter 5 degrees. Do u think it is worth to pay?

Sorry, I think this is just bull****.
Rooting has absolutely NO influence on performance whatsoever if you don´t use root access to manupulate kernel settings.
This is just a bad case of a false positive.
Just running latest Beta10 prooves this is nonsense since I have no root and same performance
Even pretty stable...each of my three attempts on Beta10 gave similar results 260tsd+.
Global Stable instead gives bad results in Antutu but shows how obsolete Antutu is because Beta10 is just a tad faster....SURE not 260tsd+ versus 160tsd. Antutu is CRAP!

Haha
DJ-Tumor said:
Sorry, I think this is just bull****.
Rooting has absolutely NO influence on performance whatsoever if you don´t use root access to manupulate kernel settings.
This is just a bad case of a false positive.
Just running latest Beta10 prooves this is nonsense since I have no root and same performance
Even pretty stable...each of my three attempts on Beta10 gave similar results 260tsd+.
Global Stable instead gives bad results in Antutu but shows how obsolete Antutu is because Beta10 is just a tad faster....SURE not 260tsd+ versus 160tsd. Antutu is CRAP!
Click to expand...
Click to collapse
I'm sure u DONT know whatever thing of rooting. Haha

running on Global MIUI 10 Beta 8.7.12 and Antutu score was 260k - but the phone doesnt feel any different from the MIUI 9 Global Stable that had "only" 170k.
not sure if performance testing is doing anything for real world usage?

I have a real benchmarking problem! Antutu, 3dmark, GFXBench crashes after random time. Trying on different FW (official global: stable miui 9, stable miui 10, dev miui 10 (android 8), dev miui 10 (android 9)) also trying xiaomi.eu frimware. Same result. CPU temperature during tests is not getting higher than 54.

Related

Skyrocket Benchmarks [12/29 - New Battery Info!]

I thought it was worth starting a thread specifically for data concerning the Skyrocket stats. For other folks posting, OC'd scores are welcome, but please also provide info on your OC frequency. Quadrant scores are pretty good random number generaters, but that's about it. If you want to post them, that's fine, but the variance is so high it's worthless trying to actually use them for reliable benchmarks. Several others are far better. (CF-Bench, Passmark, Antutu, etc.)
CF-Bench is a nice benchmark - only a 2-3% variation in the tests I've done so far, although there aren't any graphics tests unfortunately. There are several posts in this thread with CF-Bench results for comparison.
Monjori really pushes the GPU shader, is repeatable, and doesn't really depend on the CPU speeds that much. Thanks to Da_G for pointing it out.
[Update 12/29/2011] - More Battery Information (T989)
I recently purchased a battery from amazon (based on the comments that it was fully compatible with the Skyrocket.) Samsung ET-CGPK004GSTA T-Mobile SGH-T989 Battery Charger with Stand
The results have been amazing, with over 74 hours of battery life in my first test. Granted this was with light use, but the battery life has been by far consistently better than the one that my phone came with. Has anyone purchased this battery and had the same level of success? I seem to remember having to clear out the battery stats, but once I did, everything worked just fine.
Note that the serial numbers are slightly different, but all of the specifications seem to be identical otherwise. I didn't check to see if NFC was working however, it's not something I'm likely going to use.
Additional details are in post #56 of this thread, but I'll post the screenshot here. (Note that the desk charger I purchased also comes with the battery.)
[Update 12/10/2011] - Battery Information
It's been a while now since I've posted, so I'll try and clean up some of the misconceptions about the battery in the Skyrocket.
The below results are using Da_G's Kernel release 0.3. They are going to be somewhat better than stock, given that I was using both CPU profiles for Under/Over clocking to boost battery/performance respectively, but it was also a fairly high workload with an hour or so of GPS usage while driving, Performance Testing, long screen display times, and so forth.
To cut to the chase, From 100% to 10% Battery, I hit 31 hours in this test. Theoretically I could have received considerably more on another test when I had very light usage (26 hours from 100 -> 45%), I had to charge my phone for a trip the next day so I didn't finish it up.
On to the screenshots!
Total time and Utilization graph:
Time spent in each state:
End of test:
Estimated JuiceDefender Savings:
Current JuiceDefender Settings. These are hopefully close to what I used during the test, but I'd forgotten to take a screenshot. Hopefully they are close.
[Update: 11/20/2011] - Monjori scores, DA_G's OC kernel 0.4-pre.
(No screenshots here since they just show the FPS after the test.)
Stock Freq, ondemand : 47 FPS
Stock Freq, performance : 51 FPS
1809 Mhz, ondemand, 48 FPS
1809 Mhz, performance, 51 FPS
Scores from Da_G, same kernel:
Current tweaks are pushing 6700 in Antutu, 8200 in CF-Bench and 51 fps in monjori. These are currently the only benchmarks I am validating the kernel against as they are very repeatable (all tests within margin of error) unlike Quadrant!
Those scores are at 1.809GHz OC, on my 0.4pre kernel
Click to expand...
Click to collapse
[Update: 11/19/2011] - CF Bench Scores
This is using Da_G's OC kernel .3 @ 1.782 Ghz. Nice results with a linear ramp up in scores - what you'd expect with a reliable benchmark that mostly factors in CPU. I'm having some stability issues with graphics benchmarks, but there should be a fix out for it soon. I'm not going to bother with Quadrant - just glancing at the Dev thread shows wild variations in scores, exactly the same as what I saw. Gotta love those 1980's apple II style graphics though!
Another nice feature of this kernel is a stable 54 Mhz frequency! This means great battery savings.
[Update: 11/13/2011] - CF Bench Scores, Stock kernel.
{
"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"
}
[Update: 11/9/2011] - Passmark scores
I was originally going to post data from mobilebenchmark.net showing our results vs. the SGH-I777, then I looked closer and saw the charts were based on one sample. It's a good benchmark, currently in beta, and I'd encourage folks to get it from the market and submit some results. My results were 1660, or 8.2% higher than the OGS2 (based on 41 samples). I'd be interested in seeing what other people get.
This benchmark certainly has much more realistic tests than Quadrant. At least for now the SGH-I727 is one of the fastest phones available. I'm a little surprised we beat them out on the 3D graphics ranking (keep in mind this is still beta software) but at least it's a much more realistic test of what these phones are capable of doing. I'd be interested in the specific score breakdowns here to see how the CPU tests were broken down.
Main Site
Overall scores
3D Graphics Rating
Original Post:
Just wanted to get some benchmark scores and screenshots. If you have the OGS2 and would post benchmarks for comparison, I'd appreciate it. (Screenies as well so all the numbers can be compared.)
Quadrant seemed whacked to me - it was giving me scores all over the place, both low and high.
Best AnTuTu benchmark run: 5956
Some battery history. Basically around 10 hours of battery time went from 100% to 25%. It was fairly idle after 6pm, but I did some moderately heavy benchmarking at one point and had > 1hr at full clock frequency. This was largely at 50% brightness, wireless/bluetooth/gps/sync all on.
I'll grab and post more benchmarks tomorrow.
AnTuTu Benchmark over 6000! (Also a reserved post for later)
For some reason drocap shows in red.
5849
Sent from my SAMSUNG-SGH-I727 using xda premium
gharris040604 said:
For some reason drocap shows in red.
5849
Sent from my SAMSUNG-SGH-I727 using xda premium
Click to expand...
Click to collapse
I was using ScreenShot by Kastor Soft, which has worked for me with everything so far.
Skyrocket Screen comparison pictures
The OpenGL test is beautiful, and it was capping at the Vsync rate a fair amount of the time. Tons smoother than the Quadrant benchmark even though there was a huge amount more detail in the graphics. That convinces me even more that Quadrant is out of date. Unfortunately you can't get your actual results on that test without logging in to a website that didn't have a registration button, so I just skipped it for now.
So lets look at the screen display sizes and pixels. I don't have an OGS2 any more since I traded it for the Skyrocket, so for comparison I took pictures of my Captivate. (Ironically I took the captivate pictures with my Skyrocket since it was better than my old camera). The last picture was taken with my SLR camera.
One thing I saw right away was a difference if you use a downloaded font on the Skyrocket. They were crisper when seen from normal viewing distance it seemed like. Personally I'm quite happy with the screen anyway, but it did make a difference.
To my eye anyway, you can see some difference ... but only if you are so close up it's impractical. (For those of us with binocular vision anyway who read more than an inch away from the screen ... just sayin.)
Admittedly though there is some difference in vibrancy that I really can't reproduce by taking a picture with my limited equipment. I really only notice it when I switch from one phone to the other, or see them side by side.
First the Captivate. It's definitely crisp, and even the close up shot shows the very smooth letters.:
Similar picture of the Skyrocket:
Close view of Skyrocket:
Standard viewing distance:
Quadrant benchmarks
I have been very torn whether to take back my OGSGSII until I spent an
hour with the Skyrocket at the At&t store. It felt a bit laggy, the camera was
slower and the screen was definitely not as sharp. Text was fuzzy, the screen
did not seam as bright on the same settings and on lighter colored screens,
you could see the pixels. I am staying with my OGSGSII.
shootthis said:
I have been very torn whether to take back my OGSGSII until I spent an
hour with the Skyrocket at the At&t store. It felt a bit laggy, the camera was
slower and the screen was definitely not as sharp. Text was fuzzy, the screen
did not seam as bright on the same settings and on lighter colored screens,
you could see the pixels. I am staying with my OGSGSII.
Click to expand...
Click to collapse
If we don't get dev support in the next week or so this thing will go back. It's for sure laggy I agree but I love the screen on it.
Sent from my SAMSUNG-SGH-I727 using XDA App
then you probably will be taking it back. Without the source code there will be no dev support.
shootthis said:
I have been very torn whether to take back my OGSGSII until I spent an
hour with the Skyrocket at the At&t store. It felt a bit laggy, the camera was
slower and the screen was definitely not as sharp. Text was fuzzy, the screen
did not seam as bright on the same settings and on lighter colored screens,
you could see the pixels. I am staying with my OGSGSII.
Click to expand...
Click to collapse
People have been saying factory resets helped with the screen hitches. I know mine improved after removing the bloatware. For me full brightness is just too bright (on both devices) so I was usually using an Auto or 50% setting. Whether or not you like the screen is entirely subjective, and if you like the OSGS2, by all means, keep that one. (although at normal distance I'm certainly not seeing more pixellation).
I for one don't miss the Vlingo button lag!
The Quadrant scores (for me) have just been way too inconsistent for me to think it's reliable (1600 - 3500+). Here's screenshots of runs not long after you posted. (Scores at 3416, 3554, and 3710). Since it benches better than the OSGS2 scores you posted you wouldn't expect it to feel laggy. Actually this is the first time I've ran it since removing bloatware and it's faster than ever. Stock kernel and frequencies.
If we don't get dev support in the next week or so this thing will go back. It's for sure laggy I agree but I love the screen on it.
Click to expand...
Click to collapse
I'm not sure that's a realistic timeframe. Up to you, but I'd say factory reset and start removing bloatware obviously. Yeah, it's more work than flashing a ROM and have it all done for you, but to me it's part of the fun. At least I'm learning something instead of just knowing how to fire up Odin correctly.
We have CWM and root, so it's not like we can't properly use our phones for now. Even the OSGS2 took more than a week to get stable kernel's out the door and it had all the benefits of the international ROM's. But yeah, the OSGS2 will have more dev support no doubt. It's just a question of what's more important to you.
tejones36 said:
People have been saying factory resets helped with the screen hitches. I know mine improved after removing the bloatware. For me full brightness is just too bright (on both devices) so I was usually using an Auto or 50% setting. Whether or not you like the screen is entirely subjective, and if you like the OSGS2, by all means, keep that one. (although at normal distance I'm certainly not seeing more pixellation).
I for one don't miss the Vlingo button lag!
The Quadrant scores (for me) have just been way too inconsistent for me to think it's reliable (1600 - 3500+). Here's screenshots of runs not long after you posted. (Scores at 3416, 3554, and 3710). Since it benches better than the OSGS2 scores you posted you wouldn't expect it to feel laggy. Actually this is the first time I've ran it since removing bloatware and it's faster than ever. Stock kernel and frequencies.
I'm not sure that's a realistic timeframe. Up to you, but I'd say factory reset and start removing bloatware obviously. Yeah, it's more work than flashing a ROM and have it all done for you, but to me it's part of the fun. At least I'm learning something instead of just knowing how to fire up Odin correctly.
We have CWM and root, so it's not like we can't properly use our phones for now. Even the OSGS2 took more than a week to get stable kernel's out the door and it had all the benefits of the international ROM's. But yeah, the OSGS2 will have more dev support no doubt. It's just a question of what's more important to you.
Click to expand...
Click to collapse
I know all about how well this does on benchmarks, I have a post with about every app that benchmarks and the results. I have gotten rid of what I think is bloat and its helped for sure but it's still not as smooth as the original but I knew that before I bought it, I just don't know if I can deal with that lag long term... I'll figure it out in the next 20+ days. Fantastic device for sure and still head over heels better than the 4s I had for two weeks.
Sent from my SAMSUNG-SGH-I727 using XDA App
Skyler2Dope said:
I know all about how well this does on benchmarks, I have a post with about every app that benchmarks and the results. I have gotten rid of what I think is bloat and its helped for sure but it's still not as smooth as the original but I knew that before I bought it, I just don't know if I can deal with that lag long term... I'll figure it out in the next 20+ days. Fantastic device for sure and still head over heels better than the 4s I had for two weeks.
Click to expand...
Click to collapse
My benchmark's were in reply to shootthis' post, not yours ..
I recall your posts. We get the same results, but for whatever reason you are getting some kind of lag and I don't. Are you referring to that screen hitch when scrolling between screens?
Regardless, at some point the lag went away for my device. It's completely smooth, not that the slight hitch ever bothered me before. That's one reason why I can't remember when it completely went away.
I referred to your post for part of what I froze with TiBu, but there were differences. About the only other thing I can think of is I'm using setCPU profiles, and my default is the ondemand scheduler. I'm wondering if auto-brightness impacted one of my benchmarks, but probably not.
No clue. Hopefully someone will figure it out soon so we can put this thing to rest.
11/9 Update
Updated first post with Passmark benchmarks.
Horrible network speeds today - yes, I know speedtest.net isn't definitive. The main reason I ran it was that just the listing of forum posts (here) took forever to load while I was at lunch today. This was actually a faster benchmark than the server located in the city I live in ... So the poor data speed shown here was in line with what I was seeing qualitatively.
The funny thing is all of the quadrant score screenshots I posted were from the Skyrocket and not the OGSGSII.
shootthis said:
The funny thing is all of the quadrant score screenshots I posted were from the Skyrocket and not the OGSGSII.
Click to expand...
Click to collapse
Funny how? It shows that Quadrant isn't a worthwhile test. I knew that already. My scores are within a few percent of each other with Passmark, not all over the place like with Quadrant. Sort of proves the point.
First attempt with CFbench
Sent from my SAMSUNG-SGH-I727 using Tapatalk
Danielk425 said:
First attempt with CFbench
Click to expand...
Click to collapse
Nice, hadn't run across that benchmark yet. I'll pick it up!
Its a good one chain fire makes quality apps I use the pro just to support his work
Sent from my SAMSUNG-SGH-I727 using Tapatalk
Here's mine as well...
In terms of battery..., yesterday i unplugged my phone in the morning ~9am, got to work at 10am, at work I stream high quality spotify over 4G to my BT headphones practically non stop + a 1hr tele conference over my phone
low battery warning came on at ~4:30 with the phone usage actually being the biggest drain at around 55%, screen and spotify were both less than 15%
overall I'm pretty happy with that, phone almost lasts as long as my headset will while streaming unplugged
ktiedt said:
In terms of battery..., yesterday i unplugged my phone in the morning ~9am, got to work at 10am, at work I stream high quality spotify over 4G to my BT headphones practically non stop + a 1hr tele conference over my phone
low battery warning came on at ~4:30 with the phone usage actually being the biggest drain at around 55%, screen and spotify were both less than 15%
overall I'm pretty happy with that, phone almost lasts as long as my headset will while streaming unplugged
Click to expand...
Click to collapse
After monitoring all the things using battery for a couple days, I removed the "bloat". I am a very heavy data and texting user, with maybe 20 minutes of calling. Today was the first day of being bloat free and I am at 15hrs since being unplugged and I still have 68%. I'm shocked and very happy. Lets see tomorrow if it was just a fluke..

Android Multi threaded / core usage and efficiency

I have always wonder, how well dose android leverage multi core SOCS? I know from being a PC hardware geek that most developers are terrible and using all the power in multi core CPU's. I understand that in benches all the cores are used, but how often are task spread out over the cores in real life usage. My theory is that they are barely used yet I have no evidence to promote this.
Depends on the governor settings but they definitely use all cores a lot by default.
Install system monitor lite and see for yourself
{
"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"
}
Sent from my Nexus 5 using Tapatalk
rootSU said:
Depends on the governor settings but they definitely use all cores a lot by default.
Install system monitor lite and see for yourself
Sent from my Nexus 5 using Tapatalk
Click to expand...
Click to collapse
Do you think this would scale well onto the 8 core SOCS? I was just wondering if they really offered good real life performance rather than just good benchmarks
Gipjoe said:
I have always wonder, how well dose android leverage multi core SOCS? I know from being a PC hardware geek that most developers are terrible and using all the power in multi core CPU's. I understand that in benches all the cores are used, but how often are task spread out over the cores in real life usage. My theory is that they are barely used yet I have no evidence to promote this.
Click to expand...
Click to collapse
Weird. I was just wondering this myself, mainly because I just dont know what kind of specs it would take to have me use something other than my N5 and I am a phone-whore.
There aren't many tasks that need more than 2 cores active at a time. Gaming is one of the few times a 3rd core will kick in from what I've seen.
Traditionally, encoding benefits a lot from multiple cores but that isn't a task people usually attempt on their mobile.
Sent from my Nexus 5 using Tapatalk
Multi tasks
bblzd said:
There aren't many tasks that need more than 2 cores active at a time. Gaming is one of the few times a 3rd core will kick in from what I've seen.
Traditionally, encoding benefits a lot from multiple cores but that isn't a task people usually attempt on their mobile.
Sent from my Nexus 5 using Tapatalk
Click to expand...
Click to collapse
You said that it will trigger the third core while gaming. I have a Nexus 5 and it lags in a specific have game. Idon't think my Nexus is trigering all the cores as it happens on my PC with the same application. On pc All cores goes to 90% as i don't have a graphics card. On Nexus i can't get the game to run on 100% even when i drop down all the quality over performance. I know that the device can handle this game. It's just a bootle neckle a believe. It's running stock rom stock kernle On demand governor. It should trigger all the cores.....i'm very sad with that too as i have bought the device manly for use like this.
Farahjr said:
You said that it will trigger the third core while gaming. I have a Nexus 5 and it lags in a specific have game. Idon't think my Nexus is trigering all the cores as it happens on my PC with the same application. On pc All cores goes to 90% as i don't have a graphics card. On Nexus i can't get the game to run on 100% even when i drop down all the quality over performance. I know that the device can handle this game. It's just a bootle neckle a believe. It's running stock rom stock kernle On demand governor. It should trigger all the cores.....i'm very sad with that too as i have bought the device manly for use like this.
Click to expand...
Click to collapse
Which game are you talking about? Once the phone starts heating up within about 20-30minutes it will begin to thermal throttle the CPU cores and clock speeds.
bblzd said:
Which game are you talking about? Once the phone starts heating up within about 20-30minutes it will begin to thermal throttle the CPU cores and clock speeds.
Click to expand...
Click to collapse
I'm talking about Digimon Adventure (1.5GB rom) on PPSSPP it's not the only game that i'm having this problem. And i'm pretty sure that nexus is not "taking it serious". From the beggning of the game you can see it already laggin. When i zoom out and the back ground bacames large it goes around to 50% when have about 6 NPCs on the screen it hapens too. When battling its pretty unstable. Some times ti can take up to 200% and some time to 90 again on the same screen in 3 seconds just changing the perpective of the camera.
@update i have take the screenshots from system monitor.
Farahjr said:
I'm talking about Digimon Adventure (1.5GB rom) on PPSSPP it's not the only game that i'm having this problem. And i'm pretty sure that nexus is not "taking it serious". From the beggning of the game you can see it already laggin. When i zoom out and the back ground bacames large it goes around to 50% when have about 6 NPCs on the screen it hapens too. When battling its pretty unstable. Some times ti can take up to 200% and some time to 90 again onde the same screen in 3 seconds just changing the perpective of the camera.
Click to expand...
Click to collapse
That's an issue with the ROM and emulation. It's very common for emulated games to have frame rate issues and you probably need to play around with the render settings. Nothing to do with our phone specifically, check other forums for more advice on the subject.
bblzd said:
That's an issue with the ROM and emulation. It's very common for emulated games to have frame rate issues and you probably need to play around with the render settings. Nothing to do with our phone specifically, check other forums for more advice on the subject.
Click to expand...
Click to collapse
See the images above. My cpu goes only to maximium of 50% on the emulator. I have been playng other things and i feel that my device is getting some kind of botttle neck.
Sorry about the english.
Farahjr said:
See the images above. My cpu goes only to maximium of 50% on the emulator. I have been playng other things and i feel that my device is getting some kind of botttle neck.
Sorry about the english.
Click to expand...
Click to collapse
Did you down clock your CPU? If not, it's probably an issue with the emulated PSP ROM. You can play around with the emulator's render settings to try and get smoother game play but I am no expert on the subject.
bblzd said:
Did you down clock your CPU? If not, it's probably an issue with the emulated PSP ROM. You can play around with the emulator's render settings to try and get smoother game play but I am no expert on the subject.
Click to expand...
Click to collapse
Never.....my Phone is all Stock. And CPU -Z says it is in On the demand governor. The emulator does support a incrediblle variety of multi threaded process like inputs and process. These features are optional and when i turn the process multithread on, it slow down a litle more. i believe that it occurs because of the limitation of the 2 cores and the Android not allowing it to use the others. The emulator have a version for PC. The same emulator does multithread on my I3 Quad core 3.1GHz.

Antutu v6.0 is now available

Many years looking the same... but now antutu is completely redesigned and looks really nice.
You can now compare chipsets performance also ...
I know that lot's of people don't like it but many others do... I do!
Have a look... I bet you'll all like the new antutu...
I scored 69k at first test using Mokee
Just to let you know
41k on sMIUI 12.3
69K?! O.O
You make me want to try Mokee
44k on my overly crowded sMIUI. Looking at LG G4 score, we should be able to reach that point so 69k is the expected score.
leledumbo said:
44k on my overly crowded sMIUI. Looking at LG G4 score, we should be able to reach that point so 69k is the expected score.
Click to expand...
Click to collapse
your sMiui must be really messy...
got that on mine -
{
"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"
}
yorikkk said:
your sMiui must be really messy...
got that on mine -
Click to expand...
Click to collapse
Not sMIUI fault I believe, when it was still virgin I can manage 5Xk too on antutu 5. With aqua mail (3 accounts), facebook, hangouts, line, path, twitter, plurk, slack, whatsapp and many other background apps, it's hard to reach that score back.
Stable chinese MIUI rom 7.0.7.0, rooted and greenified. Managed 68500.
Sent from my Mi-4c using xda premium
66 668. I would prefer 2 points less
On sMiUI 12.10 my phone scored 65k. Forgot to take screenshot.
As States before... 69k on my Mokee... But... It is more 70k than 69... Anyway, what a machine
Ooops... 70K... Getting better! Now with a few more apps installed... So I guess it has to do with the ROM... The apps you have installed have nothing to do with antutu score i guess... Just saying...
Getting 47858 on CM12 on charger without restarting :/
Resetting and running in Airplane mode gave me 42k on charger
Running again without a charger gave me 58119...
Restarted again, ran on battery got 60070...hmmm
On charger phone got higher temps, and there is throttling.
grant_ said:
Getting 47858 on CM12 on charger without restarting :/
Resetting and running in Airplane mode gave me 42k on charger
Running again without a charger gave me 58119...
Restarted again, ran on battery got 60070...hmmm
Click to expand...
Click to collapse
Our device will get slower when it's hot due to thermal throttling, the thermal engine controls both CPU and GPU. When charging, especially when it's over 60%, the temperature will reach a high value that may exceed the threshold imposed by the thermal engine. Thus, either the core speed will be clocked down (CPU0-3 max 600, CPU4-5 max 864) or they will be totally off. Same with GPU (max 300 when hot, 600 when normal). I think Xiaomi didn't notice that many of its customers live in the equator lines where the temperature is generally warm to hot (25-35) so the threshold values will be reached much sooner than those who use the phone in norther or souther area.
It's true... But in this version of antutu the scores dont get uch a diferente after 2 or 3 tests... In previous version I used to get 57k, then 54k and then 50k... In this version I get 70k, 69k and then 69k again... Not much of a diference in results after heating... Why would this happen? I never get less than 69k
siul_lasc said:
It's true... But in this version of antutu the scores dont get uch a diferente after 2 or 3 tests... In previous version I used to get 57k, then 54k and then 50k... In this version I get 70k, 69k and then 69k again... Not much of a diference in results after heating... Why would this happen? I never get less than 69k
Click to expand...
Click to collapse
Maybe you need to put more background apps to your phone as I do
Having a screenshot of Android Assistant (both Monitoring and Processes tabs) might help the analysis.
Around 70k is a very bad score I must say, for our hardware.
On last Antutu I got 59k score and for example my friends Iphone 6s had 57k.
Now I'm getting 69k, and he is getting 120k scores.....
mrkubix said:
Around 70k is a very bad score I must say, for our hardware.
On last Antutu I got 59k score and for example my friends Iphone 6s had 57k.
Now I'm getting 69k, and he is getting 120k scores.....
Click to expand...
Click to collapse
TBH, that's not bad since you can buy FOUR Mi4C for an iPhone 6s. Twister Core in Apple A9 is the best single core processor to date (which Snapdragon 820 Kyro cores might be able to beat when it comes out seeing latest benchmark info). Our Snapdragon 808 is even nowhere near 810, how could you expect it would beat Apple A9? Here's the official AnTuTu list for each SoC:
IPhone 6s with 57k... In your dreams. Mrkubix go to sleep...
Sd820 still does not beat iPhone single core performance... Dont know how they did that twister... But they did it very well... Typhoon was already wonderful...

[Benchmark Scores] (reduced after MIUI10 update)

Hi, today after updating my Poco F1 through miui10 OTA update ,I've noticed benchmark scores are reduced marginally as compared to the score on miui9. Still I don't believe in benchmark but I found it unusual!!
On antutu I didn't found much of a difference but
On Geek bench it's way too less as compared to previous Benchmarks in miui9. I ran Geek bench for multiple times but no use!! That score came out to be way less which is not at all close to a SD845 device used to produce.
And I'm using Poco F1 6/64gb varient
*****Please post your Geek Banch scores here*****
I've included screenshots of Geek banch scores for miui9 and also miui10 for the comparison!! [And also included antutu screenshots whete I found no difference at all]
Download Benchmarks here[playstore link]
Geek bench -
https://play.google.com/store/apps/details?id=com.primatelabs.geekbench
Antutu-
https://play.google.com/store/apps/details?id=com.antutu.ABenchMark
Antutu 3d-
https://play.google.com/store/apps/details?id=com.antutu.benchmark.full
Make sure you're in a cold room. Temperature makes a huge difference
Benchmark on MIUI 10 for Poco F1 (India) - 8.10.25
I don't have AC and even then there is no heating issue that could affect benchmark results significantly, as 2-3k is normal deviation at this level of scores(i.e. 1-2%).
My antutu score improved after updating to miui 10 (8.10.25 currently) it was 265k on miui9 and it is 282k in miui 10.
As for the geekbench i have to test yet allthough i have tested 3d mark if you are interested.
I cant link to post I made on instagram when I tested it because i am new user. search akshay_rajvir user name in instagram and you will find images.
hope it is useful.
just keep in mind mate. Antutu is just a number , I also facing same problems like you but for me , The performance , battery life doesn't change at all.
But if you like to see your performance based on your scores . just try to install sphinx kernel. tell me the scores when you done.
Sent from my [device_name] using XDA-Developers Legacy app
My problem is a converse of yours. i am getting too low CPU scores...around 70k in Antutu. And almost similar in geekbench. Heavy throttling I see. Some users have said their CPU temps go higher than 90 but mine hardly goes above 60. it is throttled at 60.
Hey..... Why is it that after every update geek bench scores keep going down....... Now with 10.3.6.0 people are getting 1400 single core and 4400 multi-core thats shameful for a SD845 but on the other hand antutu score are nice and high 290k so no problem there...... So what is going on,...... And apart from that after every update internet is plagued with users review of worse battery performance...... Is this intentionally being done by MI to push newer devices ???????

Question Another battery let down

Im really unsatisfied with my s22ultra ive done everything i know or people suggested and nothing helped i think next time i go xiaomi or apple. Thats it
{
"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"
}
After looking pictures of all these threads you made about this problem it's looks like pretty normal. There is no phone in the world that doesn't drain battery if you doing something with it. Apps drain battery and games drain even more battery.
Previous threads.
Real bad battery
Im having performance issues along with battery drainage. Snapdragon 8gen1
forum.xda-developers.com
Battery sucks
Did anyone found a way to get more battery from the s22 ultra i get like 3h sot . Today it went from 85 to 60 in 40min i was playing a game but thats was really quick .. i rather buy an iphone just for the battery and performance. But i really...
forum.xda-developers.com
The problem is the SOC is inefficient because of Samsung fabs poor 4nm manufacturing node (both Exynos and Snapdragon were manufactured on this node). Nothing will help, it'll take the S23u with TSMC manufacturing the new SOCs to get good battery life.
Iyad22 said:
Im really unsatisfied with my s22ultra ive done everything i know or people suggested and nothing helped i think next time i go xiaomi or apple. Thats it
View attachment 5733863
Click to expand...
Click to collapse
If you're not satisfied with S22Ultra you won't be satisfied either with Xiaomi or any other brand, ESPECIALLY Xiaomi and their garbage OS.
Your usage seems pretty normal and the fact that you have DUAL sims and play games, your battery looks fine. NONE phone will provide drastically more power with this kind of usage.
Dayuser said:
After looking pictures of all these threads you made about this problem it's looks like pretty normal. There is no phone in the world that doesn't drain battery if you doing something with it. Apps drain battery and games drain even more battery.
Previous threads.
Real bad battery
Im having performance issues along with battery drainage. Snapdragon 8gen1
forum.xda-developers.com
Battery sucks
Did anyone found a way to get more battery from the s22 ultra i get like 3h sot . Today it went from 85 to 60 in 40min i was playing a game but thats was really quick .. i rather buy an iphone just for the battery and performance. But i really...
forum.xda-developers.com
Click to expand...
Click to collapse
I get 8 to 9h sot with the 13 pro max same or even more heavy usage i tried 60 hz 720p and i get 5hsot max
ekin_strops said:
If you're not satisfied with S22Ultra you won't be satisfied either with Xiaomi or any other brand, ESPECIALLY Xiaomi and their garbage OS.
Your usage seems pretty normal and the fact that you have DUAL sims and play games, your battery looks fine. NONE phone will provide drastically more power with this kind of usage.
Click to expand...
Click to collapse
I get 8 to 9h sot with the 13 pro max same or even more heavy usage i tried 60 hz 720p and i get 5hsot max ..... they say the 12s ultra can get alot more battery or maybe even the new pixel 7pro
Iyad22 said:
I get 8 to 9h sot with the 13 pro max same or even more heavy usage i tried 60 hz 720p and i get 5hsot max ..... they say the 12s ultra can get alot more battery or maybe even the new pixel 7pro
Click to expand...
Click to collapse
Do you use DUAL SIM in your 13 Pro Max? (Rhetorical question).
ekin_strops said:
Do you use DUAL SIM in your 13 Pro Max? (Rhetorical question).
Click to expand...
Click to collapse
Its 1 sim . Plus i used to use my s22 ultra with only one i got the new sim last month or so
I also feel like the battery it's the biggest problem of this phone . Everything else it's good for me , except for some hy temperature sometimes with it's random but mostly when it's hot outside.
I made alspicture faccurom
justkta said:
I also feel like the battery it's the biggest problem of this phone . Everything else it's good for me , except for some hy temperature sometimes with it's random but mostly when it's hot outside.
View attachment 5734497
I made alspicture faccurom
Click to expand...
Click to collapse
Watch this, it explains why. The TSMC SOCs in the s23 series should fix it but the s22 series is not good for efficiency .https://youtu.be/s0ukXDnWlTY
Thank you. Now I know the problem. But also I still don't really understand why it's taking so much battery while it's in stad by ... like the cpu should not be in use then right? I love this phone but this battery it's just awful... anyway thank you for the video. It was useful
justkta said:
Thank you. Now I know the problem. But also I still don't really understand why it's taking so much battery while it's in stad by ... like the cpu should not be in use then right? I love this phone but this battery it's just awful... anyway thank you for the video. It was useful
Click to expand...
Click to collapse
To my shock today i dont know how but just happened...... plus i never have temperature issue
Next time go Poco No use in overpaying for performance when it's not backed up by a good battery. Also... Samsung, with their well known battery issues. No, thank you. Take that advice as you wish.
Tomo123 , I don't even belive what I see there I think maybe you can make this when you let your phone at home and not using it
Iyad22 said:
To my shock today i dont know how but just happened...... plus i never have temperature issue
Click to expand...
Click to collapse
I have always goof battery with the Samsung phones at the beginning like 2 -3 mounts it last like 5-6 hours of screen on time and like 20 hours of stand by , but after this period it's just too awful maximum 3-4 hours with maximum 6-8 hours of stand by I think that it's not even making it 20 hours only with stand by but I don't know why exactly it's this happening
justkta said:
Tomo123 , I don't even belive what I see there I think maybe you can make this when you let your phone at home and not using it
Click to expand...
Click to collapse
My dear man, I really don't care if you believe it or not. It just goes to show you have no idea on how to properly optimize the phone you are using. But I am not here to spoon-feed you, learn it yourself... or live in ignorance. Your choice. Have a nice day.
The problem is the kernel not the cpu fab. A quick simple watch of the cpu frequency behavior shows this with ZERO hot plugging and to high idle frequency. I'm not talking how the old hotplug governor was. I'm talking when you turn the screen off the cpus don't hotplug and don't go under 600mhz 4 core. Very very bad for better life and simply fixing this will make the phone last loads longer for standby and adding hotpluging with lower 200mhz idle till screen touch then ramp to 600mhz will add loads of screen on time. I'm not surprised this is forgotten in a world of root is bad but back in the evo3D days you could pm power=0 and get 30 days on a hotplug governor. It's there, we just need custom kernel or Samsung (very unlikely) to fix said issues
In the S22 series, while Snapdragon variant doesn't have great batterylife, the Exynos variant is even worse.
And to improve the endurance, there are 100's of threads that will help to give you some relief. That's all.

Categories

Resources