Related
Results in a spreadsheet:
https://docs.google.com/spreadsheet/ccc?key=0AntDDKv-lS6IdGRRdDBkUkxWUVpMTVhsQk1HNUlWYUE
A little background
Supercharging is a memory management script. I believe its origin is here: http://forum.xda-developers.com/showthread.php?t=1592312
I decided to test it out to see if performance is enhanced in any measurable way. I may at some point run this thread through my UX machine to find out what people have said about it.
I used CyberGR's ROM because it has a user friendly minfree menu option and also he provided a flashable mod with the supercharging script. Also CyberGR's ROM has no issues I'm aware of, having previously ranked it high among the ICS ROMs. Other ROMs may also provide supercharging, but I'm not sure which ones.
I tested with performance/cfq at 1GHz, very clean install, apps restored from titanium backup, sync disabled, etc... same methodology as my previous ROM benchmarks.
Summary of Results
There was no significant performance difference in any configuration.
That is, no difference with or without the script, at any minfree level. The only difference I noticed was that the supercharger script defaulted to a setting of 125MB free, i.e. apps start getting killed if you have less than 125MB free. As a result, I was able to multitask fewer apps. To me, this does not seem like a benefit. (Although if you do install the script, there is nothing stopping you from adjusting your minfree to a lower level, as the options to do this are in the menu, with stick-on-reboot.)
Feel free to tell me where I am going wrong with this, as I am no expert here...
bedalus said:
Results in a spreadsheet:
https://docs.google.com/spreadsheet/ccc?key=0AntDDKv-lS6IdGRRdDBkUkxWUVpMTVhsQk1HNUlWYUE
A little background
Supercharging is a memory management script. I believe its origin is here: http://forum.xda-developers.com/showthread.php?t=1592312
I decided to test it out to see if performance is enhanced in any measurable way. I may at some point run this thread through my UX machine to find out what people have said about it.
I used CyberGR's ROM because it has a user friendly minfree menu option and also he provided a flashable mod with the supercharging script. Also CyberGR's ROM has no issues I'm aware of, having previously ranked it high among the ICS ROMs. Other ROMs may also provide supercharging, but I'm not sure which ones.
I tested with performance/cfq at 1GHz, very clean install, apps restored from titanium backup, sync disabled, etc... same methodology as my previous ROM benchmarks.
Summary of Results
There was no significant performance difference in any configuration.
That is, no difference with or without the script, at any minfree level. The only difference I noticed was that the supercharger script defaulted to a setting of 125MB free, i.e. apps start getting killed if you have less than 125MB free. As a result, I was able to multitask fewer apps. To me, this does not seem like a benefit. (Although if you do install the script, there is nothing stopping you from adjusting your minfree to a lower level, as the options to do this are in the menu, with stick-on-reboot.)
Feel free to tell me where I am going wrong with this, as I am no expert here...
Click to expand...
Click to collapse
ive been saying this for a long time. but i do feel that it could benefit older legacy devices that dont perform like newer devices(like the og droid, mytouch..)
What an amazing work!
Bedalus you're the graph's king
I respect you
I always suspected there was no real benefit to this script. Except maybe that it helps to stop launcher re draws. Although using the latest roms and kernels with CM9 bigmem does the same thing.
RcrdBrt said:
What an amazing work!
Bedalus you're the graph's king
I respect you
Click to expand...
Click to collapse
I don't want your respect. I want your obedience!
just kidding. Drunk.
bedalus said:
I don't want your respect. I want your obedience!
just kidding. Drunk.
Click to expand...
Click to collapse
youll need one of these then..
{
"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"
}
Hopefully no-one reads the quickly quoted quote below...
bedalus said:
I have one. Not kidding.
Shouldn't admit that. Drunk.
Click to expand...
Click to collapse
lmfao
oh man are you going to regret this haha
derekwilkinson said:
lmfao
oh man are you going to regret this haha
Click to expand...
Click to collapse
D'oh!
derekwilkinson said:
lmfao
oh man are you going to regret this haha
Click to expand...
Click to collapse
You should really hide it from zeppelinrox, the supercharger script creator
tchaari said:
You should really hide it from zeppelinrox, the supercharger script creator
Click to expand...
Click to collapse
Dammit... what did I miss?
heh.
Anyway if free ram is the same before and after supercharging, benchies would be usually be the same.
I'd be curious about KAK and 3G TurboCharger results tho
What I believe is similar to simms22's that this supercharger script may benefit devices older than our Nexus S, which have less RAM available.
With Nexus S and ICS, I don't think there's major problem with memory management.
It's true, though, that after multitasking with several apps simultaneously, especially those RAM-hungry apps, launcher may redraw.
But that doesn't happen anytime.
Having said that, custom kernels now provide this crazy BIGMEM hack which provides our Nexus S with around 380-400 free RAM available.
But my question is: Isn't free RAM a waste?
I learn this from my experience with iOS (please don't mock me here ).
I agree with you Glenn, it seems a waste. I see no reason why not set the minfree to the smallest amount possible, since the OS kills apps when necessary.
Yeah having alot of free memory, over a certain amount, is indeed a waste and just kills multitasking.
But having too much free ram in order to get speed is a classic case of improving one thing at the expense of another.
SuperCharger aims to keep it fast without having too much free ram.
With a higher launcher priority, everything stays snappy without having to go crazy with the free ram thing.
Thank you for your hard work in all of your benchmarks bedalus they have helped me get closer to the perfect set up.
Every time any updates come out I do a completely clean install scripts are the last thing I apply. I really can tell a difference in the feel of my device after running the V6 script. It feels much more snappy. I don't know if this would be considered a "performance gain" as it is an overall feel of my device.
Almost forgot, thank you Zeppelinrox for your awesome scripts. I run them all on my NS4G. And yes, I leave the V6 animation on
Sent from my KANG'd Nexus S 4G using Tapatalk 2
I think Android manages memory very well.
I'm a heavy multitasker.
I read a lot: news, Twitter, xda, articles in random websites using Chrome, all done while listening to music, yet I seldom get my stock launcher redraws.
This is done without a kernel having BIGMEM hack.
Compared to iOS memory management, Android is far better in this case.
I'm proud to have an Android phone among my iPhone friends.
They all always whine "Gosh! Apple really needs to give iPhone more RAM!"
And I was like "No buddy. Apple has to learn how Android manages memory."
No offense for those using iOS device.
I do also use it.
Sent from my Nexus S using xda premium
bedalus said:
I agree with you Glenn, it seems a waste. I see no reason why not set the minfree to the smallest amount possible, since the OS kills apps when necessary.
Click to expand...
Click to collapse
Lowest amount isn't wise and neither is a high amount. The listed "free" ram is used for caching and is nearly all full (use the free command in terminal for more info). Very little ram gets wasted in phone OSes. As ram gets to around 1gb+ minfree values won't see the light of day with current apps anyway.
The more cached the better, so long as it doesn't impact the currently needed portion.
Harbb said:
Lowest amount isn't wise and neither is a high amount. The listed "free" ram is used for caching and is nearly all full (use the free command in terminal for more info). Very little ram gets wasted in phone OSes. As ram gets to around 1gb+ minfree values won't see the light of day with current apps anyway.
The more cached the better, so long as it doesn't impact the currently needed portion.
Click to expand...
Click to collapse
That's exactly what I'm experiencing.
With more apps cached, multitasking is faster.
However, when I see an increase in free RAM available, that's when I "feel" my phone is snappier. But actually this gives disadvantage to multitasking. Some apps needed to be loaded from scratch.
glennkaonang said:
That's exactly what I'm experiencing.
With more apps cached, multitasking is faster.
However, when I see an increase in free RAM available, that's when I "feel" my phone is snappier. But actually this gives disadvantage to multitasking. Some apps needed to be loaded from scratch.
Click to expand...
Click to collapse
Free RAM, as listed in the [Settings > Apps > Running] section is cached RAM. None of that is "free", it is being used, however all of that is available if a currently running service requires it.
Minfrees play a role in getting rid of the "expendables" when this "free" RAM gets below a certain point (at least i'm pretty sure it is the cached RAM that is taken into account with minfrees, otherwise multitasking wouldn't be possible considering the actual unused RAM amounts). More info here. These are very useful on low memory devices, as you get above 768MB - 1GB they become near meaningless, as you'll find yourself rarely ever out of memory.
Thanks for your explanation, Harbb.
This clears any misconception I have.
Anyway I hate this thanks limit in XDA!
Having had the latest OTA popup today to update to 14.3.A.0.757 I was keen to look into it and see if it would be any better.
I did a quick Google of the firmware version and read some posts, mostly on here, about issues, the touch screen issue (which has never presented itself to me), battery usage, but most importantly to me, RAM usage.
Well I'm currently running 14.2.A.0.290 on my Z1, and have been for a while.
I have various apps installed, Facebook, Twitter, Netflix, Zedge, Flickr, Photobucket, Sko Go, and various others. I do use my phone a lot.
My phone has been running for 239 hours and some minutes, at time of posting, and is currently showing 520mb free RAM.
I note that when I got this phone new, on 4.1 I think it was, I had 0.9Gb free RAM.
Every time I update, I end up with less RAM.
Now my concern with upgrading again to Kitkat means I am worrying about RAM usage, specifically, if it uses even more, then I am going to be facing force closes and random crashes. Deja Vu.
I bought an Xperia Z, then a Z1 because my old Arc S couldn't cope with all the upgrades. My Arc S worked briliantly when I bought it, but the updates soon made it sluggish, and force closes were common place.
My Xperia Z is running fine on 4.3, but even that phone has lost RAM to the upgrades - that phone was showing 1.2Gb RAM free when I first got it, now that is down to 750mb
Updates alaways seems to equal less performance.
Has anyone compared the Antutu benchmark results before and after updates?
Installed Antutu and ran without restarting my phone, so after ten days -
{
"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'm downloading and installing the update now, so we'll see how it goes.
I'll post up the results when I'm done.
Factory resets?
Sent from my C6903 using XDA Premium 4 mobile app
Destroyedbeauty said:
Factory resets?
Sent from my C6903 using XDA Premium 4 mobile app
Click to expand...
Click to collapse
I haven't factory reset the phone, no, in fact I rarely do after running updates.
Factory resets mean having to set everything up again, accounts, settings, preferences, etc.
I have all the backup options set, and I know Google are great for maintaining my contacts, but I still have to reset Facebook integration, and re-link all of my contacts, and so on. I also like to keep my messages, so that means backing up my SMS before resetting, so it does mean some work to do a full reset.
I have never found it necessary to do a full reset.
Well the verdict is this.
Some nice improvements visually, and some extra free RAM, now seems to run with 700mb free RAM after letting the phone settle for a while.
This is an improvement, so I'm quite happy with that.
I like the unlock animation, and the fact that you can swipe in any direction now. It looks more like fairy dust, compared to the shutter effect previously.
Camera launch is now at the bottom of the lock screen, which is ok.
I had heard that the dialler app was supposed to be improved, showing favourite numbers, and pulling in info about local business on incoming calls, however, I notice no difference when accessing the app, other than the ICE (In Case of Emergency) contact at the top in the contacts listing.
I re-ran the Antutu benchmark, and it does show a DECREASE in performance
It's not much, but is is definitely a decrease.
In all honesty, the update seems to run well, seems fast, and responsive. Really good.
The notifications are grouped now so the fast links to turn off your sound, etc, are now tabbed.
I did notice that the fast links are no longer themed - they are not red, or blue, just white.
A small difference, but I like themes, and like coloured fast links, so a bit of a shame.
The icons themselves look bigger, although they take up the same amount of room.
There's probably other differences, and I'm sure I'll discover them in time.
Overall verdict - good update, impressed.
MickieH said:
Well the verdict is this.
Some nice improvements visually, and some extra free RAM, now seems to run with 700mb free RAM after letting the phone settle for a while.
This is an improvement, so I'm quite happy with that.
I like the unlock animation, and the fact that you can swipe in any direction now. It looks more like fairy dust, compared to the shutter effect previously.
Camera launch is now at the bottom of the lock screen, which is ok.
I had heard that the dialler app was supposed to be improved, showing favourite numbers, and pulling in info about local business on incoming calls, however, I notice no difference when accessing the app, other than the ICE (In Case of Emergency) contact at the top in the contacts listing.
I re-ran the Antutu benchmark, and it does show a DECREASE in performance
It's not much, but is is definitely a decrease.
In all honesty, the update seems to run well, seems fast, and responsive. Really good.
The notifications are grouped now so the fast links to turn off your sound, etc, are now tabbed.
I did notice that the fast links are no longer themed - they are not red, or blue, just white.
A small difference, but I like themes, and like coloured fast links, so a bit of a shame.
The icons themselves look bigger, although they take up the same amount of room.
There's probably other differences, and I'm sure I'll discover them in time.
Overall verdict - good update, impressed.
Click to expand...
Click to collapse
Correct me if I am wrong, but why would more free RAM necessarily be a good thing in an OS like Android? I would be happy if it would use more RAM to keep things like web browser tabs etc. in memory and improve multitasking.
I see why a memory leak is bad but otherwise I can't understand the hunt for as much free RAM as possible.
dape16 said:
Correct me if I am wrong, but why would more free RAM necessarily be a good thing in an OS like Android? I would be happy if it would use more RAM to keep things like web browser tabs etc. in memory and improve multitasking.
I see why a memory leak is bad but otherwise I can't understand the hunt for as much free RAM as possible.
Click to expand...
Click to collapse
I would be happy also to be able to run several tabs/apps and still have RAM spare.
RAM basically is the OS's workspace.
Less workspace = less room for apps to function, hence force closes and slow performace.
For example.
My old Sony Ericsson Arc S had a very small amount of RAM in comparison to todays devices, but when I bought it, it ran with over 100mb free RAM which meant I could install other apps and they would run happily. Once I upgraded the OS from Gingerbread to ICS, the amount of available RAM was less than 75mb and often 40mb, with hardly any 3rd party apps installed. This meant that the phone would look nice, running ICS, but would run slowly, and apps would force close due to insufficient RAM.
I do not want to have a great phone in the Xperia Z1 (or my Xperia Z for that matter) that will run like a bag of crap because it has not enough RAM to work in.
Imagine running the Z1 with only 100mb spare - how would it function? How fast would it be? How many force closes would the user endure? How many apps would you be able to install then??
How would you open the tabs, and apps like you say with a smaller amount of RAM?
You would try, but you would get force closes as the OS shuts down what it thinks are unnecessary processes to maintain the system.
So, this is why I hate to see RAM eaten away by updates.
That is my opinion anyway.
I understand what you are saying, and I agree in a way, I don't mind apps using RAM at all, as long as the phone has sufficient RAM to allow those apps to run alongside the system, with good speed, and no force closing.
In all honesty, I do not mind the slight reduction in the Antutu benchmark, but I do love the increase in RAM demonstrated with this update.
It's early days yet, but it's looking good.
The question is the phone uses 1 gb ram for what?!!!!
My old galaxy s2 was using just 400mb ram
And my old s3 used 500 mb ram
And my friend's galaxy s4 uses 1.1gb ram!!
What is the huge difference between s3 and s4 to use the twice amount of ram?? (this is related to the first question)
Iam running 4.4.2 now and I think 4.2.2 was better for ram consuming
My advice for you disable apps what are unnecessary for you
Enjoy!
Since KitKat is optimized for devices running with 512 MB RAM there shouldn't be any lag or force closes due to RAM usage. The RAM used on my Z1 is mostly 60-70% of the total amount but it doesn't mean that there are any slow downs. In fact it runs as smooth as it should ( a few hickups in the UI now and then but fixed with latest update) and never had I problems with multitasking yet. It would be nicer to see that the OS decreases the RAM usage but it is not a big deal as long as ii doesn't affect the performance.
Linux handle memory differently as compared to other OS.. More free ram is equal to wasted ram in Linux.. Unless you could find memory leak..
Sent from my C6903 using Tapatalk
I've seen a good amount of people complaining about lag, so i thought i'd share what i've done to get my g3 running smoothly.
* See below for kernel settings
First off I rooted my d850 with towelroot. Using titanium backup, I froze all at&t apps along with apps I don't use like kindle, uber, beats music. Also froze messages back & sync since it was always running, and I backup up my texts manually.
Everything I froze:
{
"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"
}
Now there's the whole art debate. I have noticed a big improvement in smoothness and battery life using art, but there's also compatibility issues with some apps, and xposed. When switching back to dalvik runtime you can definitely see some general UI stuttering, here's what can be done.
Kernel settings
I installed (and highly recommend) Kernel Toolkit and played around with a few settings. You'll have to install busybox too which worked fine. KEEP IN MIND we don't have a recovery to save our asses in case of bootloops so if you make any changes in trickster do so at your own risk, and when using set on boot, apply a long enough delay for you to fix any bootloops in the event that your phone doesn't like your settings.
That being said I would be surprised if anyone manages to find settings that causes reboots on the stock kernel, given that theres no voltage table to mess up, or any unique cpu governers.
Onto my settings:
I/O scheduler = cfq
Governor control (with interactive governor) -> hispeed_freq = 1267200
Here are some other changes you can make in governor control for better performance:
go_hispeed_load - change from 90 to 85
min_sample_time - change from 40000 to 50000
target_loads - change from 85 to 80 or 75 (75 = better performance)
NOTE: I have tested my settings for a few days and they seem to be stable. If you want even better performance (at the cost of some battery life) you can try some more changes DON'T use the set on boot option until you're sure they don't cause reboots! Once you deem them stable you should use set on boot with a 30 second delay. Without the set on boot option you'll have to manually reapply your settings after every bootup as kernel settings get set back to default on boot.
NOTE: I've found using the ondemand governor causes reboots.
With these settings i'm not missing art anymore. I've found that setting the scheduler to cfq takes care of most of the lag. Keep in mind that tweaking the kernel for better performance USUALLY comes at the cost of a little battery life. So you'll have to find a sweet spot that your happy with in terms of performance and battery life. Feel free to discuss and share your kernel settings
I'm probably the only one experiencing more lag with ART than not. With ART enabled my G3 was stuttering everywhere and I had to pull the battery ~3 times in 2 hours because the screen just wouldn't turn on.
It's silky smooth on dalvik though, lol
KiNG OMaR said:
I'm probably the only one experiencing more lag with ART than not. With ART enabled my G3 was stuttering everywhere and I had to pull the battery ~3 times in 2 hours because the screen just wouldn't turn on.
It's silky smooth on dalvik though, lol
Click to expand...
Click to collapse
Yeah sounds like compatibility issues with some of your apps.
Tested ondemand governor, caused reboots.
Surely a third party launcher will just sort all lag issues out?
Why would it do that? Of course it helps with launcher lag but not with lag occuring in menus etc.
Oh and thanks to Heisenberg! Tried your settings and with art enabled everything works nice and smoothly. No more stuttering.
Please let us know how stable these settings end up being. Thanks!
I think people are getting confused over what is lag and what is not.
Lag is when the UI is delayed when it shouldn't be - this can be caused by a number of factors including wrong cpu clock speeds, overheating or lack of quality ram management (usually due to one or two poorly coded apps eating everything up!)
What most of you are experiencing on Dalvik is not lag, its just the way the JIT compiler works, apps on Dalvik will take a split second to open while it compiles what it needs, generally speaking this is not noticable, but we all know people on XDA always notice everything....
If you are experiencing lag then it is most likely caused by something other than Dalvik the ART build on this device is outdated doesn't offer anything (seriously!) over Dalvik, this in no way reflects what we will see on Android L.
My g3 suffers terribly from redraw it just should not be happening on something this powerful
Lennyuk said:
I think people are getting confused over what is lag and what is not.
Lag is when the UI is delayed when it shouldn't be - this can be caused by a number of factors including wrong cpu clock speeds, overheating or lack of quality ram management (usually due to one or two poorly coded apps eating everything up!)
What most of you are experiencing on Dalvik is not lag, its just the way the JIT compiler works, apps on Dalvik will take a split second to open while it compiles what it needs, generally speaking this is not noticable, but we all know people on XDA always notice everything....
If you are experiencing lag then it is most likely caused by something other than Dalvik the ART build on this device is outdated doesn't offer anything (seriously!) over Dalvik, this in no way reflects what we will see on Android L.
Click to expand...
Click to collapse
I had some noticable stuttering in many actions like in menus and scrolling. After changing to ART and enabling these tweaks they are gone. Placebo or not I'm happy with the results, no stutters or lag of whatever you want to call it I'm aware the ART included is nothing like ART on L, but to me it just helps that little bit.
Lookinen said:
I had some noticable stuttering in many actions like in menus and scrolling. After changing to ART and enabling these tweaks they are gone. Placebo or not I'm happy with the results, no stutters or lag of whatever you want to call it I'm aware the ART included is nothing like ART on L, but to me it just helps that little bit.
Click to expand...
Click to collapse
That's placebo.
ART only affects lag, that is the time between pressing on an icon and the app opening. ART pre compiles the bytecode needed to run an app beforehand thus eliminating the need to do it on the fly which dalvik does (it should also improve battery life as converting it to bytecode takes cpu cycles). It doesn't do anything for stutter or scrolling. By yeah the ART in 4.4.2 is old code as well and not optimized yet so there will be little improvement in lag also.
So be it. It doesn't matter to me if its placebo or not. Feels nicer to use and that is what matters.
Still not getting the ART love. For milli seconds of improvement you lose 25% of storage space. Not seeing the net positive at all, but a big a$$ net negative.
rushless said:
Still not getting the ART love. For milli seconds of improvement you lose 25% of storage space. Not seeing the net positive at all, but a big a$$ net negative.
Click to expand...
Click to collapse
You'll have no choice when "L" is released
Batfink33 said:
You'll have no choice when "L" is released
Click to expand...
Click to collapse
I will miss Xposed if they can't get it to work right on ART
I'll try your guide. Lag and redraw is not what I prefer in a new flagship. Art is more than placebo, it's notable better but I'll like to use Xposed.
Batfink33 said:
You'll have no choice when "L" is released
Click to expand...
Click to collapse
SadLy true
Seems to do some difference. Even with art I had some lag in app drawer and it seems to be gone. Placebo or not, it's feels a bit more snappy in app drawer so that's another one to beat the dust ? So far I'll keep art on, I have loads of space and launcher and apps are notable snappier with it in. Its not like my M8 but it's a bit closer.
art
Sorry for a very noobish question, but can someone tell me what art is? I got my lg g3 last week and quiet playing a bit with it. I got the international version 3/32gb (LG-D855 V10a) and wanted to root it to somehow boost my battery life. Its fine and my battery can last a day with my normal usage and im very happy with it (waaaaaay much better than my ip5), but if i can squeeze more juice and make it last a day and a half or even more, im willing to root and take full pledge over my device.
I also experience having error that my device is heating up and won't be able to max out brightness, weird coz when it happen im actually doing nothing with my device, i just adjust brightness and it wont let me to max it out, so as of now im considering it as a bug and hoping that an update will fix it. It does warm up a bit but not the same as my toasty galaxy note 2 and galaxy s4 so i can live with it, but if i can remove the energy being waste and boost battery life by rooting it, ill forget OTA.
Anyway, back to my main concern,,, what is art? Hue hue
Hey mate. Art is android run time. Do a quick google search on how to change it or one of the pro's on here can go into the finer details on the benefits of using it.
hey guys
this is a thread for all benchmark results. i heard of 35k in antutu.... :silly:
dunno what i have to write more ^^ my results come when my tab is arrived
sry for my bad english^^
{
"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"
}
this is from HP's version of the note 7 the Slate 7 Extreme 4400us
EVGA Tegra Note 7 Stock 4.3 stock kernel.
antutu bench: 35732
UK Advent Vega Tegra Note 7
Evga Tegra Note 7
Evga Tegra note 7 - stock 4.2 in max performance mode..
Thoughts after 20 tests.....
Ran antutu 20 times... All were done in max performance mode.. Test ran minimum of 20% faster when plugged in to ac adapter. Scored as low as 28000 when battery was under 30% (not plugged in). Plugged in avg score was a little over 36000. Any one else seeing the same results?
Hi guys,
I don't want to be too rude with you but look what you can get from your device :
Note : I use Shaky156's marvelous kernel)
From left to right : Antutu 5, Antutu 5, 3DMark, PC-Mark
chp2 said:
Hi guys,
I don't want to be too rude with you but look what you can get from your device :
Note : I use Shaky156's marvelous kernel)
From left to right : Antutu 5, Antutu 5, 3DMark, PC-Mark
Click to expand...
Click to collapse
I have to be honest, when running Shaky's kernel, I see a lot more stuttering and hangups in games vs the stock kernel, not to mention the extra drain on battery life. I put mine back to stock after a week or so and it's been running better, so benchmarks be damned
rchrdcrg said:
I have to be honest, when running Shaky's kernel, I see a lot more stuttering and hangups in games vs the stock kernel, not to mention the extra drain on battery life. I put mine back to stock after a week or so and it's been running better, so benchmarks be damned
Click to expand...
Click to collapse
Could be normal sice not all the devices are born equal.
Concerning the battery drain, For my personal usage it's more than enough (I play PSX emu, Gaming, Videos, music, surfing, ...) and I don't complain till I have power for whole day.
For sure there are also lot of apps wich can cause wakelocks and don't let the device to go in deepsleep when needed. (Facebook, FB Messenger, Skype,...). I don't use them.
chp2 said:
Hi guys,
I don't want to be too rude with you but look what you can get from your device :
Note : I use Shaky156's marvelous kernel)
From left to right : Antutu 5, Antutu 5, 3DMark, PC-Mark
Click to expand...
Click to collapse
how the f*** did you do that? mine is 39635 on antutu 5 (also using the kernel by shaky)
are you using the odexed or de-odexed stock rom? any display settings changed? (for example: color correction Native or sRGB; Tegra Prism disabled?)
are you using any tweaks beside that? (like editing .prop or .xml files)
GodOfPsychos said:
how the f*** did you do that? mine is 39635 on antutu 5 (also using the kernel by shaky)
are you using the odexed or de-odexed stock rom? any display settings changed? (for example: color correction Native or sRGB; Tegra Prism disabled?)
are you using any tweaks beside that? (like editing .prop or .xml files)
Click to expand...
Click to collapse
First of all, be carefull with benchs since it's not representative to real life use.
Antutu scores are very unconsistent depending alot of background services and apps. Moreover, personnaly, I don't trust Antutu anymore. My pictures were there just to show them for comparaison.
I've done these bench just after upgrading my kernel to latest Shaky's kernel and after I removed (deactivated) lot of unecesary services and apps (specially these communicating apps). I use (the fantastic) LBE Security Master for that.
The main word is : the device has to be clean during the benchs.
For Info : I use Stock ROM whith TWRP (I think it's Odexed). Everything else is just standard :
- Color correction : Native
- Prism should be enabled by default since I don't seen how to disable in the params.
- Enable GPU acceleration in the Developper menu.
- IO : FIOPS
- Governor : USERSPACE
As a side note, try to put your device in deep sleep prior a test. Shaky's OC Kernel has a modded throtling system and the benchs results are better if the device is cold (this is true with every device I own).
chp2 said:
First of all, be carefull with benchs since it's not representative to real life use.
Antutu scores are very unconsistent depending alot of background services and apps. Moreover, personnaly, I don't trust Antutu anymore. My pictures were there just to show them for comparaison.
I've done these bench just after upgrading my kernel to latest Shaky's kernel and after I removed (deactivated) lot of unecesary services and apps (specially these communicating apps). I use (the fantastic) LBE Security Master for that.
The main word is : the device has to be clean during the benchs.
For Info : I use Stock ROM whith TWRP (I think it's Odexed). Everything else is just standard :
- Color correction : Native
- Prism should be enabled by default since I don't seen how to disable in the params.
- Enable GPU acceleration in the Developper menu.
- IO : FIOPS
- Governor : USERSPACE
As a side note, try to put your device in deep sleep prior a test. Shaky's OC Kernel has a modded throtling system and the benchs results are better if the device is cold (this is true with every device I own).
Click to expand...
Click to collapse
i know that antutu scores are not speaking for the every day usage but i love the "theoretical" overall performance you can read out from that score.
btw you mentioned you removed unnecessary apps and services. i already debloated mine a bit (just left the apps which i am not sure if they are safe or unsafe to delete). and for the rest of the apps (like facebook, whatsapp etc) i am using greenify (since this is alot better than nsaver in my opinion because it has more options) to hibernate them.
could you mention which apps and services you deleted/deactivated and if they are safe or unsafe to delete/deactivate like those bunch of tegra services that are running in backround while not really needed? for example, do i really need the SHIELD audio service running the whole time?
btw i got my tegra note 7 two days ago. i am just curious how much we can modify and tweak it to its full potencial
GodOfPsychos said:
i know that antutu scores are not speaking for the every day usage but i love the "theoretical" overall performance you can read out from that score.
btw you mentioned you removed unnecessary apps and services. i already debloated mine a bit (just left the apps which i am not sure if they are safe or unsafe to delete). and for the rest of the apps (like facebook, whatsapp etc) i am using greenify (since this is alot better than nsaver in my opinion because it has more options) to hibernate them.
could you mention which apps and services you deleted/deactivated and if they are safe or unsafe to delete/deactivate like those bunch of tegra services that are running in backround while not really needed? for example, do i really need the SHIELD audio service running the whole time?
btw i got my tegra note 7 two days ago. i am just curious how much we can modify and tweak it to its full potencial
Click to expand...
Click to collapse
I can't realy list all the services and apps I disabled because it depend on your personal taste.. And there are a lot to disable.... (all these non usefull Google apps first and non latin keyboard inputs). Do a little search.on each service/app can help.
For Nvidia apps, exept Tegra Zone, all the others are enabled
chp2 said:
I can't realy list all the services and apps I disabled because it depend on your personal taste.. And there are a lot to disable.... (all these non usefull Google apps first and non latin keyboard inputs). Do a little search.on each service/app can help.
For Nvidia apps, exept Tegra Zone, all the others are enabled
Click to expand...
Click to collapse
Ok. Well tegrazone is a nice app especially for me because it lists graphic intense or optimized games. I left just a few google apps like google + (for beta registration communitys), maps, play games and the google keyboard( updated to latest version with material design).
Btw do you have a screenshot of your antutue score with all the details (which process got how much points like "cpu integer" or "cpu float" etc)
Oh besides that, do you also have a problem with the lates youtube app? The "famous" server connection error on every video you want to watch? If yes, is there already a fix for that?
Sent from my TegraNote-P1640 using XDA Free mobile app
GodOfPsychos said:
Ok. Well tegrazone is a nice app especially for me because it lists graphic intense or optimized games. I left just a few google apps like google + (for beta registration communitys), maps, play games and the google keyboard( updated to latest version with material design).
Btw do you have a screenshot of your antutue score with all the details (which process got how much points like "cpu integer" or "cpu float" etc)
Oh besides that, do you also have a problem with the lates youtube app? The "famous" server connection error on every video you want to watch? If yes, is there already a fix for that?
Sent from my TegraNote-P1640 using XDA Free mobile app
Click to expand...
Click to collapse
Concerning all these apps you like, at least, set them as "manual start". Don't let them start automaticaly!
- First, because they use lot of memory and cpu for nothing. (Antutu bench gives beter results with more availaible RAM)
- Second, because most of them are always trying to communicate with external servers. (creating more cpu interrupts)
- Third, some of them don't let your device to go to deep sleep and thus consume battery.
- Fourth, because of personnal privacy.
Just Install "LBE Security Master" (and the xposed translation pack) just to see how much they consume.... how many "autostarts" without real reason.... how many are accessing your personnal data and communicating that data.
Then adjust on your likes.
Concerning Antutu, I uninstalled it some time ago and now, I don't think it will be relevant since all my personnal apps are installed.
Ok here is my Benchmarks.
Advent Tegra Note 7 Running 4.4.2 GMS OTA 2.5 with my Modded LBH Partition flashed via APX.
3D Mark with Autu
On Stock Android 5.1 Lollipop
Hello and welcome!
This morning I began wandering the XDA ZUK Z2 forums, looking at all these new threads and posts. I'm really impressed with all the progress for the Z2 that developers have been making lately. So I decided to try a few new ROMs.
I have always found really interesting to test new ROMs, set them up, install everything back and then use the ROM for a few days to see how everything goes... Only to find out that a new feature has been released or a new version of the ROM I'm currently using and the process begins again, it's really cool :victory:
So today I picked a few ROMs, and decided to try them all, run a few benchs and write down the results. At the end I thought I could share the data gathered and perhaps we could get some idea of how these ROMs perform.
So the general process was the next:
Wipe everything but internal.
Flash ROM and ROM alone, no GAPPs, Magisk or MODs.
Wipe Dalvik and Cache (dev recommendation).
Shutdown from recovery, boot, install benchmark apps and restart.
Clear recents and begin benchmarks.
Notes:
TWRP used was latest official version.
Baseband is the latest official baseband for the chinese Z2.
ROM version was the latest as of yesterday.
Kept the phone above 40% battery, so battery charging was needed between ROM flashing from time to time.
Cooled phone between benchmarks, as to prevent throttling from affecting the benchmarks, however didn't add cooling when benchmarking.
Alright then! Let's get down to it!
Data:
{
"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"
}
SSE: Sling Shot Extreme.
SSE scores: Overall scrore, graphics score, physics score. Ditto for Vulkan.
Geekbench 4 scores: Single core, multi core.
Geekbench 4 COM: Compute benchmark.
Now let's talk a bit about these numbers. As you can see Oreo ROMs get higher scores in both 3DMark and Geekbench benchmarks, with Cosmic getting the higher score in the Sling Shot Extreme OpenGL benchmark and Cypher in the Vulkan variant, although not by much. In Geekbench Jaguar leads with the higher score in both the single core and multi core test. Now look at every Oreo multi core score, it is higher than its Nougat counterpart, I wonder if this is in part because of the drivers or OS optimizations, however I also wonder how this shows, if it show at all, in real world usage.
I would like to note that if I'm not mistaken Jaguar had increased CPU frequencies, not sure about the GPU though, and that could be one factor for the increased scores, however OMNI got really close to Jaguar and again, if I'm not mistaken, Jaguar is based on OMNI. I would like to point that as the day passed the room temperature increased (hot day), so take the temp thresholds with a grain of salt. The battery percentage loss found in the table is that which 3DMark reported in the OpenGL test.
One thing is that even when the numbers varies from ROM to ROM the difference is not that huge (around 12% max in the multi core CPU test), and most importantly... how this translates to real world usage?
So perhaps you noticed the last column, Dolphin. I decided to install the latest Dolphin version and run The Legend of Zelda - Wind Waker intro screen and write down the highest and lowest FPS reported by the emulator. What we find is that the scores do not always reflect into the emulator FPS. Also there is a 'DoT' in almost all ROMs, what does that means? Well DoT stands for slowDown on Touch, yeah it makes no sense but it stuck. When running the emulator if the phone detects no input from the touch screen the FPS slow down by approximately 40%, that's what DoT means, I had to keep touching the screen to get full FPS, only on Citrus this was not the case, and I was getting full FPS with no input whatsoever.
So what all this means? Surely we always like to see higher numbers, I know do. But with time I have found that higher numbers are not necessarily better, or that it does not translate to better performance, battery life or SoT, stability, features, etc., for example in Oreo ROMs I always liked more how colors showed in the screen, even though we have KCAL in Nougat ROMs too. Also I ran these benchmarks only once, so they are not one hundred percent accurate. Usually more runs = more score accuracy, so take these with a grain of salt.
For fun I decided to OC the CPU when using Citrus 3.5, set Adreno boost to high (honestly don't know what it does just that it boosts something related to the GPU I guess haha) and even when the benchmarks didn't improve that much (actually it got a worse multi core CPU score) I got better Dolphin FPS, and that just proves my point.
And that was just one real world test which I think it's heavier on the CPU. I would like to test an intensive GPU game and get some data on that. Also, numbers don't say much when talking about customization and features, so to really feel a ROM you have to test it out yourself.
It would have been awesome if I had have the time to test more ROMs or kernels, I just picked some ROMs from the Telegram group and some here randomly. But! That means that there is more out there for you to try out! As I said in a post this morning: That just means you have a lot of options to choose from and have fun installing, customizing and using the ROMs in your phone.
So yeah, had fun doing this and wanted to share. A HUGE thanks to all the developers out there! Thank you for giving life to our device! :highfive:
EDIT. Please forgive my english, I'm still learning
Great info , thanks :fingers-crossed:
Cypher OS Oreo download link please!
sanketp60 said:
Cypher OS Oreo download link please!
Click to expand...
Click to collapse
Sure! It has the 'Now Playing' feature, really neat.
Download
Thanks to the Lenovo ZUK Z2 - Portal Telegram group for the link.
Avan7 said:
Sure! It has the 'Now Playing' feature, really neat.
Download
Thanks to the Lenovo ZUK Z2 - Portal Telegram group for the link.
Click to expand...
Click to collapse
Any bug around there in that rom?
suggestion for best gaming and battery friendly rom.
kashyapp said:
Any bug around there in that rom?
Click to expand...
Click to collapse
Actually I just installed and ran the benchmarks right away, so I couldn't really say about bugs or share an overall experience.
amit gaikwad said:
suggestion for best gaming and battery friendly rom.
Click to expand...
Click to collapse
I think the most battery friendly ROM would be stock. However custom ROMs keep improving, so that's awesome. Other than that, this was just an experiment to get some numbers therefore I can't really say about battery.
There are so many factors, like battery drain on light and moderate usage and heavy CPU and/or GPU usage, deep sleep, hpereat, stability, user experiencie, etc.
It is hard to measure these with benchmarks alone.
Fascinating post, thank you for doing this!! I just bought this device for 7k refurb (approx 100$), should easily last a year or more.
Would it be possible to do a run with the Pixel Experience rom? Or is that one similar to the other ones you've already done? Just doing my research on this device and the various roms, its blowing me away how much horsepower exists in this thing for the low price!
Whats Citrus 3.5 OC?
Link Anyone?
Thanks for the comparison. But I was wondering, AEX is arguably the most popular ROM for Z2, so why skip that one?
frozenfirevr said:
Thanks for the comparison. But I was wondering, AEX is arguably the most popular ROM for Z2, so why skip that one?
Click to expand...
Click to collapse
+1
Sent from my Z2 Plus using XDA Labs
sanketp60 said:
Whats Citrus 3.5 OC?
Link Anyone?
Click to expand...
Click to collapse
Citrus 3.5 by miju but with an OC applied with the custom Kernel Adiutor for the Z2
frozenfirevr said:
Thanks for the comparison. But I was wondering, AEX is arguably the most popular ROM for Z2, so why skip that one?
Click to expand...
Click to collapse
I understand and the reason is that I decided to try a batch of ROMs I hadn't tried before. I wanted to try many more, but time didn't allow it. Perhaps sometime close I'll do another run :fingers-crossed:
Very interesting! Thanks for your time.
Im considering to flash jaguar... im now on aex 5.2 but seems that jaguar is smoother and run good. Waiting for new latest builds to test this two and decide.
Avan7 said:
I understand and the reason is that I decided to try a batch of ROMs I hadn't tried before. I wanted to try many more, but time didn't allow it. Perhaps sometime close I'll do another run :fingers-crossed:
Click to expand...
Click to collapse
Cool, thanks. Mind sharing your experience with each of the ROM also, if you spent any time with them? Have you chosen any one?