I know, I'm sorry, benchmarks, who needs them? The reason I'm testing is because I bought this phone used and I want to make sure that it's not broken in some way. I have a friend who is flipping HTC 10's and he's been doing benchmarks on each one that comes across his desk, and it seems like they're quite a bit different than mine.
On the stock rom my Antutu score has ranged from 108,000 to 152,000. Each time there's up to a 20,000 point difference. In my experience, your benchmark scores shouldn't change much unless you change something, so why the 45K range in scores? Not only that, but in Lineage 14.1 my scores drop significantly, down to 72,000, with the highest at 98,000.
I performed a RUU today and my scores are still flexing between 108 and 118K.
GeekBench 4 is a similar issue, my multicore performance has ranged from 1700 to 4100, and my single core from 1100 to 1700.
On a regular basis my Droid turbo (from 2014, running Lineage 14.1) is beating my HTC One in both tests. Do I have a bad phone, or is this normal behavior for the Sprint 10?
My buddy gets about 4000/1700 in geekbench and about 118K in Antutu, and his scores don't seem to change very much. He's running Lineage 14.1.
thunder2132 said:
I know, I'm sorry, benchmarks, who needs them? The reason I'm testing is because I bought this phone used and I want to make sure that it's not broken in some way. I have a friend who is flipping HTC 10's and he's been doing benchmarks on each one that comes across his desk, and it seems like they're quite a bit different than mine.
On the stock rom my Antutu score has ranged from 108,000 to 152,000. Each time there's up to a 20,000 point difference. In my experience, your benchmark scores shouldn't change much unless you change something, so why the 45K range in scores? Not only that, but in Lineage 14.1 my scores drop significantly, down to 72,000, with the highest at 98,000.
I performed a RUU today and my scores are still flexing between 108 and 118K.
GeekBench 4 is a similar issue, my multicore performance has ranged from 1700 to 4100, and my single core from 1100 to 1700.
On a regular basis my Droid turbo (from 2014, running Lineage 14.1) is beating my HTC One in both tests. Do I have a bad phone, or is this normal behavior for the Sprint 10?
My buddy gets about 4000/1700 in geekbench and about 118K in Antutu, and his scores don't seem to change very much. He's running Lineage 14.1.
Click to expand...
Click to collapse
Perhaps it's thermal throttling?
Nougat update improved things on that front. Are you on Marshmallow? Also you can try going into developer option and setting High Performance mode and then doing your test to see if you get improved scores.
Tarima said:
Perhaps it's thermal throttling?
Nougat update improved things on that front. Are you on Marshmallow? Also you can try going into developer option and setting High Performance mode and then doing your test to see if you get improved scores.
Click to expand...
Click to collapse
I think you're probably right on the thermal throttling side of things. I'm on Lineage 14.1 now, and if I run a benchmark after the phone has been idle for a while it's in the 138K range where it's supposed to be. Overclocked on stock is where I'm getting closer to 145K. I think I was doing too many tests, and when they were bad, I'd flash to another ROM (while charging) which didn't help with the thermal side of things.
Thermal throttling also would explain why if you do one test, then immediately repeat it the phone will score worse, which happens every time I try it.
I'm not on MM, I did a RUU to the latest version of N and then unlocked from there.
Related
As we all know quadrant is no reliable measure for speed. At least I knew this for a while now and it was repeated and quoted many times.
This article tells anybody with a functioning brain (that is used of course) that quadrant means pretty much nothing.
I can't help to run it from time to time anyway
So I sat on the to... in my room in front of my computer with my phone. I9000 with supersonic ROM and the remount script from adrenaline shot 7. I sat there and said to myself "how hight can you score in quadrant LOL"
I started quadrant up and ran the benchmark: 2309
Then I opened the task manager-> Exit all & Clear memory
Then via long press homebutton back to quadrant to run the benchmark again score: 2453
But since I am a programmer and can imagine all kinds of optimizations and caching I pressed the back button and just ran it again just after it finished
Score: 2675
How the hell could anyone call that a benchmark?^^
just to be sure could anyone confirm that behavior? And does anyone know of a mor reliable alternative? I'd like to collect that knowledge in this thread.
TL;DR: quadrant sucks, you know anything better or want to flame away: do it here
Those are not the actual numbers from my first experiment, I repeated the scenario just now and took the numbers from those runs.
Additional runs scored 2775, 2907 and 2820, that's just silly
I think this behaviour is well known and has to do with JIT optimizations or something like that
allotrios said:
I think this behaviour is well known and has to do with JIT optimizations or something like that
Click to expand...
Click to collapse
The reason is irrelevant. The fact it doesn't provide a reliable benchmark is.
no benchmark is precise if you don't use it as intended. Quadrant produces a reliable comparative benchmark when used as designed: run it five times, remove the lowest and highest scores and average the remaining 3 -- that is your benchmark. You may not like it, but that is how it is designed to be used.
Now if you want to be pedantic, you could reasonably test again, by running quadrant 5 times, removing the outliers and average your 3 remaining scores. Repeat 10 times and then tell me how your average scores do or do not vary: they will in fact be within a narrow range, your actual benchmark.
Alternatively, tell us which benchmark produces the same score each run, as that appears to be the sum total of your objection to quadrant.
There are other benchmarks, such as Caffiene Mark, AnTuTu and NenaMark, but they are all apps just as Quadrant is and all require several runs and averaging to produce a comparable benchmark.
Moreover, the primary use of any benchmark is to compare firmware (kernel and rom) builds on the same phone to see relative performance gain and drop.
A benchmark is supposed to give way of comparing the capabilities of a given device. This means that a device with a high average score implies a better device than a lower score.
But the Quadrant score does nothing of this sort! In a competition with a friend I achieved an average Quadrant score of about 4300, with a peak of 4462. According to Quadrant my device is a lot better than the OP! Which is just not true.
Quadrant is unreliable as a benchmark, no matter how it is "designed to be used".
Sent from my GT-I9000 using XDA App
whaave said:
But the Quadrant score does nothing of this sort! In a competition with a friend I achieved an average Quadrant score of about 4300, with a peak of 4462. According to Quadrant my device is a lot better than the OP! Which is just not true.
Click to expand...
Click to collapse
You're doing it wrong.
lgsshedden said:
Moreover, the primary use of any benchmark is to compare firmware (kernel and rom) builds on the same phone to see relative performance gain and drop.
Click to expand...
Click to collapse
Quadrant scores are useless. I've used custom roms with scores of 2500+ but they aren't as smooth as stock roms, which only have scores of 1600-1800.
Antutu is indeed quite reliable imho. My results never fluctuate more than +-5% on the same config. That's an acceptable range, considering I don't set cpu governor to performance before running my tests.
Sent from my GT-I9000 using XDA App
upichie said:
You're doing it wrong.
Click to expand...
Click to collapse
w00t?
Quadrant does not reflect performance, and therefore can not be used as a comparison parameter.
It can't be much worse than I thought.
My phone with 2.1 and 'lag fix' scored 2200 and lagged so bad I wanted to throw it against a wall multiple times a day.
With stock 2.3 quadrant can be ~1000 but the phone runs much smoother.
Other than the obvious file systems I/O 'cheats' that resulted in the above, there is also the frame rate cap that makes the GPU tests useless as well.
if your trying to measure height with a scale , u wont get your answer .
The only benchmark tool that ever reflected how the phone felt in my hands , in real life usage is linpack .
changing OC / kernel is mainly the only thing that will affect linpack if your trying to use it to compare roms ill efer you to my first statement .
In order to have a good feel of a rom / set up on the phone , use some apps that will use lots of ressources , for example TW4 launcher , go in there scroll a lot open gallery (if you have many pics) scroll thru them and repeat ... Any benchmark tools will basically tell you the 'ability of your device ' ( comparing 2 different models like an inspire and an sgs2 for example will be accurate )
ZioGTS said:
Antutu is indeed quite reliable imho. My results never fluctuate more than +-5% on the same config. That's an acceptable range, considering I don't set cpu governor to performance before running my tests.
Click to expand...
Click to collapse
I recently tried Passmark Mobile. Still a beta version, but I like it. Test results reflect real performance improvement and degradation pretty closely, particularly for what concerns I/O and memory speed.
I've recently upgraded to the ARHD 5.0.0 Rom and my Quadrant score has dropped from an average of 4800 down to an average of 3800.
I realise quadrant is not the most reliable benchmark for realworld performance but that seems like a pretty big drop.
Anyone else know what could have happened?
Thanks.
What has happened is that it's meaningless Quadrant score has dropped. I'd just ignore it altogether.
Solaris81 said:
I've recently upgraded to the ARHD 5.0.0 Rom and my Quadrant score has dropped from an average of 4800 down to an average of 3800.
I realise quadrant is not the most reliable benchmark for realworld performance but that seems like a pretty big drop.
Anyone else know what could have happened?
Thanks.
Click to expand...
Click to collapse
People are reporting lower AnTuTu scores after 1.29 so if the ROM you're using is 1.29 based what you're observing could be a result of the update. Since Nvidia's code and drivers are proprietary the devs use it pretty much as-is.
Read this...
http://forum.xda-developers.com/showpost.php?p=25601709&postcount=626
And this...
http://forum.xda-developers.com/showthread.php?t=1632354
As Nvidia plays around with low-level code and drivers to address battery life and fix the graphics issues they'll be tweaking the behaviour of the CPU. There were some pretty odd changes to the CPU code in the Prime while Nvidia was tackling its launch issues.
You have to do quadrant in stock rom.
On 1.26 and 1.28 stock rom it's easily around 4,600 to 4,800. Should be around the same for stock 1.29. I'll see it when I get the Asia ota.
I've tried most roms around and some of their versions are low and some after their update becomes on par with stock in terms of quadrant. It's due to the fact that they are doing heavy tweaking on the rom thus affecting performance a bit (whether for better or worse) . Good thing is they can easily rectify the problem on their next releases.
Although benchmark is not a clear indication of real world performance but it's always nice to get high scores nonetheless. Lol.
Sent from my HTC One X using xda premium
Not sure if it's just me but Geekbench results for multi-core have dropped quite significantly after the final Nougat release.
With Marshmallow it scored around 4400, with Nougat it's around 3200 which is quite a big drop.
Anyone else getting the same?
i got 3039 for pixel c and my nexus 9 got 3340. both with interactive governor. i would guess if the govenor was some performance type it would register higher numbers.
dkryder said:
i got 3039 for pixel c and my nexus 9 got 3340. both with interactive governor. i would guess if the govenor was some performance type it would register higher numbers.
Click to expand...
Click to collapse
Most devices are set to interactive anyway, that wouldn't result in such a large drop in multi-core performance.
Good to know it's not just my PIxel C.
1408 single / 3138 multi here :/
I have stuttering when scrolling on Chrome, but i didnt have this problem on MM and Remix OS
gtaadicto92 said:
1408 single / 3138 multi here :/
I have stuttering when scrolling on Chrome, but i didnt have this problem on MM and Remix OS
Click to expand...
Click to collapse
Thanks for testing.
I'm assuming this isn't intentional from Google.
Same here
I'm on preview 5 and I have similar geekbench results. I am running a rooted pixel c and got similar results with the performance governor as well.
Shocky2 said:
Not sure if it's just me but Geekbench results for multi-core have dropped quite significantly after the final Nougat release.
With Marshmallow it scored around 4400, with Nougat it's around 3200 which is quite a big drop.
Anyone else getting the same?
Click to expand...
Click to collapse
Maybe GeekBench isn't tuned properly for Nougat.
brando56894 said:
Maybe GeekBench isn't tuned properly for Nougat.
Click to expand...
Click to collapse
Seeing the last update for it was in Feb 2016, I would say this ^ makes sense.
I just tried out photo mate r3, which has a benchmark function as well. That is a raw photo editing app. In Marshmallow the benchmark takes 33 seconds. In nougat it takes at least 45 seconds. Very similar to the drop in performance seen in Geekbench. Unless photo mate and Geekbench are doing the exact same things, this seems like a performance regression.
It got even worse with Geekbench 4. I am now clocking 1200 single core and 2400 multi-core on Nougat. The individual stats are pretty pathetic compared to other devices. I wonder what happened?
Shocky2 said:
Not sure if it's just me but Geekbench results for multi-core have dropped quite significantly after the final Nougat release.
With Marshmallow it scored around 4400, with Nougat it's around 3200 which is quite a big drop.
Anyone else getting the same?
Click to expand...
Click to collapse
Read the reviews. Other people are having the same issues with these benchmark apps showing lower scores that are inconsistent. I don't think the Pixel C is by itself in these crappy scores.
bluestang said:
Read the reviews. Other people are having the same issues with these benchmark apps showing lower scores that are inconsistent. I don't think the Pixel C is by itself in these crappy scores.
Click to expand...
Click to collapse
Yeah. But I still get a ten second difference between Marshmallow and nougat in photo mate. I now get the same result as I do on my LG G4 (with its significantly worse processor). Nothing has changed except nougat. That's at least two apps where nougat performance is significantly worse than Marshmallow on the pixel c.
tim.malone said:
Yeah. But I still get a ten second difference between Marshmallow and nougat in photo mate. I now get the same result as I do on my LG G4 (with its significantly worse processor). Nothing has changed except nougat. That's at least two apps where nougat performance is significantly worse than Marshmallow on the pixel c.
Click to expand...
Click to collapse
Load marshmallow.
Sent from my Pixel C using Tapatalk
Thanks for the constructive reply. Next time I experience a problem in Android I will be sure to ignore it.
I think he may be saying the benchmarks FUBARed and just enjoy Nougat???
hey, i run geekbench 4 with the cpu bench and i got a really good score.
but with the compute bench (graphics) i got a really worse result.
i got 1216 points, the nexus 9 with the older soc gets 3689 points.
I think the problem is the graphics driver. some news pages says, android nougat will use a new graphic interface with the latest opengl version.
maybe nvidia hat not yet released a new driver and the older one can´t be use with the full power.
I also noticed this. I found another renderscript (the API Google uses for running computations on the GPU) benchmark and it show similarly bad results on pixel c. I also found this issue https://code.google.com/p/android/issues/detail?id=204712
That indicates that the pixel c didn't ship with a Nvidia specific renderscript driver, which might be why the performance is bad. Just a guess, but I don't think it is even using the GPU. Which is a shame. The pixel c should fly in this test. I wonder if it's possible to load in the driver from the shield TV?
ravn83 said:
hey, i run geekbench 4 with the cpu bench and i got a really good score.
but with the compute bench (graphics) i got a really worse result.
i got 1216 points, the nexus 9 with the older soc gets 3689 points.
I think the problem is the graphics driver. some news pages says, android nougat will use a new graphic interface with the latest opengl version.
maybe nvidia hat not yet released a new driver and the older one can´t be use with the full power.
Click to expand...
Click to collapse
I started one for g4 plus now for g5 plus .
Cosmic os 2.1 unofficial
Elemental x kernel over clocked
What benchmark program are you using?
username8611 said:
What benchmark program are you using?
Click to expand...
Click to collapse
Antutu
PureNexus using ElementalX stock CPU speeds and GPU governor, CFQ, custom CPU governor settings
Lineage OMS with ElementalX kernel stock CPU speed and governor. ZEN with custom readahead.
This is kind of useless, benchmark comparison means nothing if it is not on the same device with same set of apps installed.
Sent from my LG G5 using XDA Labs
suhridkhan said:
This is kind of useless, benchmark comparison means nothing if it is not on the same device with same set of apps installed.
Sent from my LG G5 using XDA Labs
Click to expand...
Click to collapse
That doesn't make any sense. Devices are manufactured to a certain tolerance and winning the "silicon lottery" doesn't make a device faster, it makes it more overclockable. Device to device, stock for stock, the difference should be at most a few thousand points from each other. It should be pretty obvious to kill all background apps and processes before benchmarking so apps installed don't matter either. If Facebook is too important to kill for 10 minutes then that person shouldn't worry about benchmarking.
Device to device are obviously going to vary. But a varience of 10k+ points is a pretty good indicator of one set up running slightly better than the other and it's interesting to compare what is the most optimized settings. I can play with my CPU governor all day and get repeatable results +/- 500 - 1000 points. Both me and my wife had a Nexus 5 and with identical settings we both benchmarked very similar. To say it is a useless test is ignorant. If people look at this as a pissing match to see who's "better" then yeah, I see this being a dumb and useless thread. But I think most people who do this want to know what settings, ROM, and kernel are best optimized for performance.
Edit: https://www.phonearena.com/phones/Motorola-Moto-G5-Plus_id10398/benchmarks
63,191
https://www.youtube.com/watch?v=345eKlssdH8
62,769
http://www.fonearena.com/blog/214719/moto-g5-plus-review.html
62,893
https://www.pcmag.com/review/352573/motorola-moto-g5-plus
63,845
http://www.guidingtech.com/65986/moto-g5-plus-vs-redmi-note-4/
62,896
5 different devices, all tested stock within right around 1,000 points of each other.
username8611 said:
That doesn't make any sense. Devices are manufactured to a certain tolerance and winning the "silicon lottery" doesn't make a device faster, it makes it more overclockable. Device to device, stock for stock, the difference should be at most a few thousand points from each other. It should be pretty obvious to kill all background apps and processes before benchmarking so apps installed don't matter either. If Facebook is too important to kill for 10 minutes then that person shouldn't worry about benchmarking.
Device to device are obviously going to vary. But a varience of 10k+ points is a pretty good indicator of one set up running slightly better than the other and it's interesting to compare what is the most optimized settings. I can play with my CPU governor all day and get repeatable results +/- 500 - 1000 points. Both me and my wife had a Nexus 5 and with identical settings we both benchmarked very similar. To say it is a useless test is ignorant. If people look at this as a pissing match to see who's "better" then yeah, I see this being a dumb and useless thread. But I think most people who do this want to know what settings, ROM, and kernel are best optimized for performance.
Edit: https://www.phonearena.com/phones/Motorola-Moto-G5-Plus_id10398/benchmarks
63,191
https://www.youtube.com/watch?v=345eKlssdH8
62,769
http://www.fonearena.com/blog/214719/moto-g5-plus-review.html
62,893
https://www.pcmag.com/review/352573/motorola-moto-g5-plus
63,845
http://www.guidingtech.com/65986/moto-g5-plus-vs-redmi-note-4/
62,896
5 different devices, all tested stock within right around 1,000 points of each other.
Click to expand...
Click to collapse
Thank you for taking the time to write a long response. But, I believe you may have just proved my point. I believe the test results of different roms should be well within 'around 1,000 points of each other'. Unless-
a. the rom is very poorly optimized - score would be lower.
b. the kernel is overclocked - score could be slightly higher.
c. user error (lots of background apps).
suhridkhan said:
Thank you for taking the time to write a long response. But, I believe you may have just proved my point. I believe the test results of different roms should be well within 'around 1,000 points of each other'. Unless-
a. the rom is very poorly optimized - score would be lower.
b. the kernel is overclocked - score could be slightly higher.
c. user error (lots of background apps).
Click to expand...
Click to collapse
I really don't know how else to explain this to you. OP got a lower score than me, yet is overclocked. So it stands to reason that either "a. the rom is very poorly optimized - score would be lower" or "b. the kernel is overclocked - score could be slightly higher" or "c. user error (lots of background apps)" is the reason for it. But wait, the performance should be slightly higher for an overclock except that it isn't. That's the whole reason to benchmark. Another possibility is that since I've heard ElementalX is currently having overclock issues, it may be reverting to its nominal frequency, which I believe is 1.4Ghz. How would this person have known that if not for comparing benchmarks? According to you, they can't compare to stock benchmarks because it's a different set of apps installed and a different ROM and in fact can't compare it to anyone because it's a different device, albeit the same model.
Benchmarks show performance differences, regardless of whether or not they are large enough to even notice on a day to day basis. It shows technical differences and if you think technical differences mean jack squat, then why are you even commenting in this thread? It's the same theory when you throw a car on a dyno. You're going to notice small differences between each run, but when you have two of the same model cars with the same engine, and one consistently puts out 30HP more than the other, there's probably a reason for it.
To reiterate what I said in my first reply, for people who want to compare optimization between different ROMs, kernels, and technical settings such as CPU governors and schedulers, benchmarking is not useless. Not in this method of testing and not across identical devices with different software. The baseline or "stock vs stock" comparison shows that the benchmark is measuring with an adequate amount of accuracy and that multiple devices in stock form are performing equally before being modified. Just because it doesn't mean anything to you doesn't mean that it means nothing at all.
I did some research and things like backround apps running in airplane mode scripts like lightning blade. all these things make a difference. I was running kernel over clocked in interactive mode with lightning script. If I set to performance my score was significantly higher I was hoping this would give users a better way to set up and optimize their device not to compare roms running same device. Yes at first I thought about that then realized it wouldn't make a lot of sense. Im hoping some of u guys will hop on board and help test kernel roms and other mods so maybe we can get the best out of our device thanks guys.
Guys I am switching Rom for best battery life and performance .
viperos latest build antutu around 110000
Jaguar rom little bit more
same with other rom like crdorid citric aex etc
i did test with stock and custom kernel also but same result I tried jaguar heliox
even geekbench are very low single core is 17hundred and multi-score is around 2900
Help me asap
Single core geekbench at 1700 is not low, and high benchmark score doesnt do anything rather than bragging on the internet, in that case, just stole another result that is high already
but other showing more
pipyakas said:
Single core geekbench at 1700 is not low, and high benchmark score doesnt do anything rather than bragging on the internet, in that case, just stole another result that is high already
Click to expand...
Click to collapse
i also have lemax2 which have 4gb ram and 32gb internal and snapdragon 820. And in lemax 2 i got 1700 single core and 4000 multicore.
amit gaikwad said:
i also have lemax2 which have 4gb ram and 32gb internal and snapdragon 820. And in lemax 2 i got 1700 single core and 4000 multicore.
Click to expand...
Click to collapse
so what? it mostly due to overhead in the software and thermal config that the cpu cannot run all cores at max speed for the entire time benchmarking. if you want high benchmark score then dont use this phone apparently
If you still insisted on getting highest benchmark score, get jaguar kernel, put everything to best performance and highest thermal limit, then throw the phone into a fridge. Maybe it would get better