Develop Bugs

ext2 does help! score 2221 + NAND INTACT *EDITED 21AUG 2018HKT

I know that many people worry about frequent reads/writes will damage the NAND, so I began a new test days before without using the NAND
this time, I switched from rfs to ext2, with some mods to the kernel to load the ext2 partitions, no OC
note that this is only serve as a real case to show u ext2's performance and to prove that rfs is one of the problem areas. u guys can implement it your ways with further improvements, perhaps with ext3, ext4, yaffs, etc
EDIT: 21AUG 2018HKT
- some ppl said that it is easy to cheat the score obtained in quadrant by moving around the quadrant's data files. yes, i agree that. but it wil be much easier to edit it in photoshxp for even 30k score. and although some ppl said the score itself is meaningless, which i agree with it also, but it provides a mean to let u know how ur phone is performed. it is not an indicator for whether ur videos or sounds can be played smoothly since it depends on a lot of things as well
- also note that if u formatted the /sdcard in different file format, some of the recovery mode functions wont work since it will only mount rfs/vfat, so u have to modify the stock recovery or rom manager's recovery manually, or by mounting the partitions via shell
{
"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"
}

so this will lead to a whole new type of lagfix?

So this is the base for the ultimate lagfix?

daijirok said:
so this will lead to a whole new type of lagfix?
Click to expand...
Click to collapse
well, I dono. let you guys decide

Nice to see the score split up in the different parts. It shows that the inital I/O of GS wasn't really that bad. Also i guess we can expect about 750 points of CPU score with Froyo (looking at the Nexus), so maybe a 3000 score is within reach .

toca79 said:
So this is the base for the ultimate lagfix?
Click to expand...
Click to collapse
i dono think so, more to come i think

mbr01 said:
Nice to see the score split up in the different parts. It shows that the inital I/O of GS wasn't really that bad. Also i guess we can expect about 750 points of CPU score with Froyo (looking at the Nexus), so maybe a 3000 score is within reach .
Click to expand...
Click to collapse
right, i saw one with 27xx or something already with old tricks. so 3000 should not be a problem with 2.2

ykk_five said:
i dono think so, more to come i think
Click to expand...
Click to collapse
Great work anyway!!

ykk_five said:
right, i saw one with 27xx or something already with old tricks. so 3000 should not be a problem with 2.2
Click to expand...
Click to collapse
ykk_five, first, many thanks for the tip. The score of 2711 was done during your first initial suggestion, where the fix was temporary. In additional, I happen to have the overclock kernel (by raspdeep) installed. Hence, I just ran it to see how much mileage we can get from the SGS.
Having said that, RyanZA and Chainfire had implemented excellent lagfixes. With your latest test, I'm really not sure how much more improvement we can get. Many of us are now very happy with what they have, and I am not sure if there are anymore juice to be squeezed out from this.
Let's hear it from the more experienced users, and perhaps this latest discovery might be as exciting as your first.
Greetings from Singapore.

g00ndu said:
ykk_five, first, many thanks for the tip. The score of 2711 was done during your first initial suggestion, where the fix was temporary. In additional, I happen to have the overclock kernel (by raspdeep) installed. Hence, I just ran it to see how much mileage we can get from the SGS.
Having said that, RyanZA and Chainfire had implemented excellent lagfixes. With your latest test, I'm really not sure how much more improvement we can get. Many of us are now very happy with what they have, and I am not sure if there are anymore juice to be squeezed out from this.
Let's hear it from the more experienced users, and perhaps this latest discovery might be as exciting as your first.
Greetings from Singapore.
Click to expand...
Click to collapse
thx a lot, i really hope that u guys will enjoy it!

I have a consistent score of 990+ without any lag fix on a froyo ROM. Previously I used a lag fixed JM5 which score 2000+, however with everyday use the froyo ROM seems snappier. Good to see all efforts to speed things up and eliminate the lag, but I'm not sure higher benchmarks score translate to a better user experience.
What I'm hoping for now is a fix than enables smooth scrolling of contacts & call logs seen on the desire.
Good work though!

Frostfree said:
I have a consistent score of 990+ without any lag fix on a froyo ROM. Previously I used a lag fixed JM5 which score 2000+, however with everyday use the froyo ROM seems snappier. Good to see all efforts to speed things up and eliminate the lag, but I'm not sure higher benchmarks score translate to a better user experience.
What I'm hoping for now is a fix than enables smooth scrolling of contacts & call logs seen on the desire.
Good work though!
Click to expand...
Click to collapse
u can compare the IO stat, although it's a roughly comparison

g00ndu said:
ykk_five, first, many thanks for the tip. The score of 2711 was done during your first initial suggestion, where the fix was temporary. In additional, I happen to have the overclock kernel (by raspdeep) installed. Hence, I just ran it to see how much mileage we can get from the SGS.
Having said that, RyanZA and Chainfire had implemented excellent lagfixes. With your latest test, I'm really not sure how much more improvement we can get. Many of us are now very happy with what they have, and I am not sure if there are anymore juice to be squeezed out from this.
Let's hear it from the more experienced users, and perhaps this latest discovery might be as exciting as your first.
Greetings from Singapore.
Click to expand...
Click to collapse
This kind of solution should be a lot better as it doesn't use any space (or so i undertand).

pls read my newly added comments at the 1st page
EDITED 21AUG 2018HKT

So what exactly did you do to get these new scores? Did you format the /data partition as EXT2 to get these scores?

I understand someone may not like to spend time to answer noob's questions. But what is the point to show just the Quadrant score and then tell others you have done something in this Development board? Should it be presented in a more technical way?

Yes please give some more details on what exactly you did.
Sent from my GT-I9000 using Tapatalk

I wonder if Froyo will bump this score even higher.
I just performed a test on the HTC Desire, with newly installed Froyo (2.2). On the scale above, it scores between the Nexus and the Droid, but closer to the Droid. What intrigued me is that the reason the score was lower than the Nexus, was the CPU score. The other colours were about the same as the Nexus One. The CPU part was still bigger than anything else (up to the 1/4 of the green part on the droid below it) but about 4/5 that of the Nexus. Sorry I couldn't post a screenshot.
Which is really weird, since the two devices share the same CPU. (are the reference scores accurate?)
EDIT: I tried running it again, and I got a higher CPU score this time, so now it almost matches the Nexus (up to 1/2 of the green part on the Droid) Overall score 1301. So scratch the above comment, I guess the reference scores are the highest you can get.
Now in the picture of the Galaxy above, the biggest part is the IO score, which is logical because of the modification, but the CPU score is very small. So assuming that Froyo will bump up the cpu part like on the Nexus and the Desire, we'll be looking one hell of a record score, right?
Has anyone with a JP firmware tried this lagfix?

The better CPU scores on FROYO are due to the JIT Compiller! The JP firmwares Have JIT but it is not active still! Or atleast not to the point it is on the N1! I hope that the final firmware will have full speed.

@ykk_five - I'm not really sure what NEW point you're making, but I love that you're still working on this!
Didn't we pretty much already know that rfs was the appaent problem with the i/o? You proved that early on as far as I can tell. Paul O and mimocan assumed it was slow hardware with their fixes, but you showed that a different fs fixed it on the same memory space. Some of us have been clamoring for a reformat of the internal rom space as the solution since you first posted your findings.
What I find more interesting is that the i/o on the stock SGS in Quadrant isn't that bad - it appears comparable to most of the other phones, except the Motorola Droid X! I wonder why it's so fast on that one.
My own earlier READ-speed testing done with reading a 200mb data file from the command line with the dd command showed that read speeds on RFS internal /data partition vs ext4 on sdcard partition were pretty close. Seems like RFS can serve up the reads pretty darned fast (Class 6 speeds or better). Well over 13Mb/s from my post over on Modaco:
Code:
BEST SPEED ON EXTERNAL SDCARD EXT 4 = 15.79 MB/S
BEST SPEED ON EXTERNAL SDCARD FAT32 = 15.6 MB/S
BEST SPEED ON INTERNAL SDCARD FAT32 = 13.1 MB/S
BEST SPEED ON INTERNAL ROM AREA = 13.6 MB/S
Since we're now doing the loop mount ext2 on top of RFS fix, I thought I should test the read speeds there. I ran four tests, they ranged from 1.1MB/s to 8.1MB/s:
Code:
BEST SPEED ON LOOP MOUNTED EXT2 /DATA/DATA = 8.1MB/S
That's right, it's 50% SLOWER to read files off the ext 2 /data/data partition (with RyanZ's 2-3 method) than just raw RFS on the best read, 1/13th of the speed in some cases!
Seems hard to imagine how this fix speeds things up other than cacheing writes, which shouldn't impact the opening of files, unless the lag in the first place was some sort of delay reading the files while other things were written to the disk.
With that in mind, apparently the Quadrant speedup is in the WRITE part of the i/o tests, wouldn't you agree? That would be because of the cached writes ext2 does vs RFS?
If this is the case, then high Quadrant scores aside, none of these fixes should really do much to improve the normal look and feel of using the phone, as you alluded to with your comment about playing back videos/music. RFS or ext will serve up the data at more than we need it.
It makes me wonder why so many of us assert that the phone is indeed much snappier on opening apps and stuff. I'm sure it's not a placebo for me, the phone definitely feels faster with any of the lag fixes we use.
What we really need is to see a breakdown in Quadrant's i/o benchmark that distinguishes between the read and writes.
I really don't like that with all these fixes, no one really knows why they speed the phone up.
I'm tempted to go back to just stock JM5 and see what happens.

FASTEST desire rom?

im looking for the fastest rom on the desire
the recently hacked xperia x10 that has the same hardware as the desire but with a smaller ram has scored on quadrant 1882 running 2.1 and a reported 2558 on the soon to be released 2.2
so what im looking for is a way to blow that score...the highest iv scored so far on my desire is 1460 with 2.2 vanilla stock roms(AOSP, CM, OD) and an overclocked cpu with the performance setting
so is there any other roms that can score as the xperia?

As far as i know the MIUI roms score the highest in quadrant, but is there any real reason for wanting a high score?

mmk92 said:
im looking for the fastest rom on the desire
the recently hacked xperia x10 that has the same hardware as the desire but with a smaller ram has scored on quadrant 1882 running 2.1 and a reported 2558 on the soon to be released 2.2
so what im looking for is a way to blow that score...the highest iv scored so far on my desire is 1460 with 2.2 vanilla stock roms(AOSP, CM, OD) and an overclocked cpu with the performance setting
so is there any other roms that can score as the xperia?
Click to expand...
Click to collapse
So you need a trick...to surpass a hack that makes a bad device, made by a bad manifacturer,running an outdated os version on bad hardware with laughable support and mind numbingly low representation to compete vs a such legenday device as desire???????????????????????????
step 1 Flash ANY aosp rom
step 2 use smartass governor on 250-1xxx frequency....
good..now u can proove to ur friend that bought a device born dead that rs is better than his.... /clap

mmk92 said:
im looking for the fastest rom on the desire
the recently hacked xperia x10 that has the same hardware as the desire but with a smaller ram has scored on quadrant 1882 running 2.1 and a reported 2558 on the soon to be released 2.2
so what im looking for is a way to blow that score...the highest iv scored so far on my desire is 1460 with 2.2 vanilla stock roms(AOSP, CM, OD) and an overclocked cpu with the performance setting
so is there any other roms that can score as the xperia?
Click to expand...
Click to collapse
my cousins galaxy s vibrant on a custom rom scored 2200+ on multiple occasions.
IMHO i like defrost, it feels fast and smooth w/ launcher pro and what not.

Defrost 6.0e, o'c to 1190. Quadrant 1900+ (performance governor).

People have posted Quadrant scores of 1800-1950 with overclocked Desires. The processor is decent but rumour has it that the score is not as high as it could be because the Desire's ROM is a bit slow. Also, results vary wildly. It's not uncommon for one test to be in the 1300s and the next in the 1800s. If you want to post a high score, be prepared to run a lot of tests.. or to use photoshop like everybody else.
You'll want an AOSP Kernel for overclocking and the highest speed scores have been posted by people with MIUI Roms (but that could just be because those Roms immitate the iPhone and thus appeal to the more status conscious, ePeen swinging crowd).
Sure the Desire streams 720p H264 video without hiccups on it's small screen
It plays 3D games with a decent enough framerate on it's rather uncomfortably small screen
But erm.. it's a bloody phone. Seriously... what do you want from it?
Test your new code for calculating the 798426th decimal of pi?
Calculating AES private keys?

The fastest roms are definitely OpenDesire 4.0.36 and Cyanogenmod 6.1 RC1, both equipped with some of the latest vorkKernels: you'll often get over 2000 on Quadrant (for what it's worth) with those. Also Oxygen 0.2.2 is pretty fast, but not as fast as them. Defrost is ok too but it's not as stable as those imho. Same story with MIUI...

i got 1990 using latest stock htc rom, richard tripps v5 kernel, and the I/O mod. im going to try for 2000 on my personal cooked rom 2moz and take a pic

soory if false.. but is not big quadrant score doesnt guarantee the phone / roms perform well ?

AndroHero said:
i got 1990 using latest stock htc rom, richard tripps v5 kernel, and the I/O mod. im going to try for 2000 on my personal cooked rom 2moz and take a pic
Click to expand...
Click to collapse
Woot?!? How do you get such huge scores on a sense rom?

vnvman said:
Woot?!? How do you get such huge scores on a sense rom?
Click to expand...
Click to collapse
by bind mounting /data/ onto a 2gig ext4 partition, it massively increases the I/O score on quadrant, its not suggested for low class sd cards though as it will slow your phone down loads, it was sibere who developed the method and script in this thread, http://forum.xda-developers.com/showpost.php?p=9362929&postcount=24 and dosent work with apps2sd, but that isnt needed as your internal storage will become your ext, so for example i have 2GB of inernal storage on my desire

OxygeN is the fastest/most responsive I have used. It is still technically in beta but I have been using it since day one with a smile.
Very slick, light and great battery life.
Previous fave was opendesire (OxygeN developer AdamG's previous work)

AndroHero said:
by bind mounting /data/ onto a 2gig ext4 partition, it massively increases the I/O score on quadrant, its not suggested for low class sd cards though as it will slow your phone down loads, it was sibere who developed the method and script in this thread, http://forum.xda-developers.com/showpost.php?p=9362929&postcount=24 and dosent work with apps2sd, but that isnt needed as your internal storage will become your ext, so for example i have 2GB of inernal storage on my desire
Click to expand...
Click to collapse
Lol sounds great, i wonder how i missed that
I guess i'll try that as soon as i get a new sdcard, currently with my 8gb one i can't lose all that space, i need it for my music

martijn_bakker said:
People have posted Quadrant scores of 1800-1950 with overclocked Desires. The processor is decent but rumour has it that the score is not as high as it could be because the Desire's ROM is a bit slow. Also, results vary wildly. It's not uncommon for one test to be in the 1300s and the next in the 1800s. If you want to post a high score, be prepared to run a lot of tests.. or to use photoshop like everybody else.
You'll want an AOSP Kernel for overclocking and the highest speed scores have been posted by people with MIUI Roms (but that could just be because those Roms immitate the iPhone and thus appeal to the more status conscious, ePeen swinging crowd).
Sure the Desire streams 720p H264 video without hiccups on it's small screen
It plays 3D games with a decent enough framerate on it's rather uncomfortably small screen
But erm.. it's a bloody phone. Seriously... what do you want from it?
Test your new code for calculating the 798426th decimal of pi?
Calculating AES private keys?
Click to expand...
Click to collapse
some flash videos don't run that well as on optimised overclocked roms.

I'm running the latest LeeDroid ROM which I love - but when I play Reckless Racing it stutters a lot. Perhaps its my class 2 memory card?

I have installed several ROMs. Now I am using LeedDroid. I think, that it is best balanced between speed and battery life.

You guys seem pretty knowledgeable, so here goes... How can I get the fastest quadrant scores possible, mainly to smoke my friends galaxy s, what can I do (ie: roms, kernels, partitioning... also disadvantages of each)?. ATM, I am getting average 1600's, @1.19GHz (v3 froyo kernel or something), smartass govenor, Leedroid v2.2f, and a APPS2SD enabled class 2 2GB samsung (300mb ext3). I will be getting a Patriot class 10 8gb soon, and I would like some tips of what are the fastest roms (stable is preferred), best kernels and things like hacking the I/O to increase scores. Thanks in advance.

NickB95 said:
You guys seem pretty knowledgeable, so here goes... How can I get the fastest quadrant scores possible, mainly to smoke my friends galaxy s, what can I do (ie: roms, kernels, partitioning... also disadvantages of each)?. ATM, I am getting average 1600's, @1.19GHz (v3 froyo kernel or something), smartass govenor, Leedroid v2.2f, and a APPS2SD enabled class 2 2GB samsung (300mb ext3). I will be getting a Patriot class 10 8gb soon, and I would like some tips of what are the fastest roms (stable is preferred), best kernels and things like hacking the I/O to increase scores. Thanks in advance.
Click to expand...
Click to collapse
I suggest you go the data2ext path with either rxmixhd v14 or ownhereDHD v1.6 ...

NickB95 said:
You guys seem pretty knowledgeable, so here goes... How can I get the fastest quadrant scores possible, mainly to smoke my friends galaxy s, what can I do (ie: roms, kernels, partitioning... also disadvantages of each)?. ATM, I am getting average 1600's, @1.19GHz (v3 froyo kernel or something), smartass govenor, Leedroid v2.2f, and a APPS2SD enabled class 2 2GB samsung (300mb ext3). I will be getting a Patriot class 10 8gb soon, and I would like some tips of what are the fastest roms (stable is preferred), best kernels and things like hacking the I/O to increase scores. Thanks in advance.
Click to expand...
Click to collapse
Check that up

It's silly going for big Quadrant scores, it means nothing.
This is my MIUI_Au ROM, test build (not public release)
Desire @ stock speed (998Mhz). I haven't bothered testing it @ 1.19Ghz yet, but I reckon it'll smash 4k (not that it means anything)
{
"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"
}

ROM Benchmark Scores

I got bored tonight and tested a few of the popular and recent ROM's for this phone. Below is the method of testing and the results.
I took these steps for -every- ROM tested.
Plugged phone in via USB to my computer and left it plugged in.
Factory reset using clockwork
format /system as well (i've seen a few rom's suggest this)
Installed ROM
Booted into ROM and let sit for 5 minutes without touching it
Stepped through the initial setup without setting up a google account (if applicable to the ROM)
Reboot
Let sit for 1 min
Installed AnTuTu Benchmark v2.6 from sdcard.
Ran AnTuTu Benchmark using default settings and selecting /sdcard/external_sd as the sdcard
This is NOT ment to imply one ROM is better than another. It's only intention is to share datapoints. Some of the ROM's are beta as well, so their final release may be different.
My intention is to show how ROMs perform based on the AnTuTu benchmark. I wish to use a ROM that is quick and efficient. Others may appreciate more eye-candy at the cost of performance. To each his own. So I hope some find this data useful.
Android Zombie v4.0 (7272)
{
"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"
}
Foxstar beta v1.0 (5689)
Revenge Of Macnut (5778)
Tuesday|Medieval|Midnight| (7136)
DARKSIDE:IGITAL::WARFARE v2.1 (5771)
Dewizzed-V2.5.2 (7045)
And just for kicks, here's my phone, running normally, fully loaded.
* Long time lurker, first time poster since I figured this may be a way to contribute to the forum.
* Also not here to promote any ROM. I'm unaffilliated.

Thanks!
Sent from my SGH-T989 using xda premium

Uh you know foxstar is the only one you ran at stock speeds? The rest are at 1700. Might want to re run that one.
Also you would have to run each test three to five times to get any sort of meaningful results bc the scores will fluctuate.
Sent from my SGH-T989 using XDA App

About the clock speeds, I did that on purpose. The purpose was to show the performance of the ROM as the creators intended.
Do you think running it multiple times will make a huge difference? I figured so long as I did the same steps for all ROMs it should reflect a decent standard.
Sent from my SGH-T989 using XDA App

matthewdavis said:
About the clock speeds, I did that on purpose. The purpose was to show the performance of the ROM as the creators intended.
Do you think running it multiple times will make a huge difference? I figured so long as I did the same steps for all ROMs it should reflect a decent standard.
Sent from my SGH-T989 using XDA App
Click to expand...
Click to collapse
You did that on purpose? Interesting. You might want to put in bold the speeds you are running each test on. Seemed like a mistake if you ask me. All of those other roms do not come with default cpu max at 1700MHz.
Also, yes running them multiple times will make a HUGE difference. I have done it when bench marking graphics cards, cpu overclocks and with my phone. Try it....you will be very surprised.

G1ForFun said:
You did that on purpose? Interesting. You might want to put in bold the speeds you are running each test on. Seemed like a mistake if you ask me. All of those other roms do not come with default cpu max at 1700MHz.
Also, yes running them multiple times will make a HUGE difference. I have done it when bench marking graphics cards, cpu overclocks and with my phone. Try it....you will be very surprised.
Click to expand...
Click to collapse
Echoing this, Even if it's defaulted to a lower clock speed, you can't leave it that to provide an accurate data set. You also need a large spread, run hem 5 or 10 times and average them.
Even if benchmarks are a little silly in my book, still some interesting data.

G1ForFun said:
Uh you know foxstar is the only one you ran at stock speeds? The rest are at 1700. Might want to re run that one.
Also you would have to run each test three to five times to get any sort of meaningful results bc the scores will fluctuate.
Sent from my SGH-T989 using XDA App
Click to expand...
Click to collapse
Not true. Darkside Warfare was run at 1350.

IMO I think it's pointless unless you clock them at the same speed I mean you have digital warfare under clocked at 1.3 -_- might also want to mention which kernel each 1 is using and at the end of the day it's not about how each 1 scores but how it does in real life use I remember when I used the jugs kernel clocked at 1.89 I believe it scores over 8k easy but would lag on scrolling and other things but faux kernel would score a little over 7k and would run like butter clocked at 1.7. Now I'm using digital warfare with the kernel it comes with and I think it run incredible with speed, smoothness, and battery with good scores
Sent from my SGH-T989 using Tapatalk

mavblues said:
Not true. Darkside Warfare was run at 1350.
Click to expand...
Click to collapse
Mistake...
I was checking the screens from xda app. Your right.
Regardless...the roms should all be ran at one uniform speed for the best comparison.
Sent from my SGH-T989 using XDA App

avarize said:
Echoing this, Even if it's defaulted to a lower clock speed, you can't leave it that to provide an accurate data set. You also need a large spread, run hem 5 or 10 times and average them.
Even if benchmarks are a little silly in my book, still some interesting data.
Click to expand...
Click to collapse
This is true, however with an asterisk. Running them subsequently will 100% boost scores, mostly because of CPU stepping. Continually running them keeps them clocked at full speeds more often, especially in the beginning.
Also, same clocks please? Use setCPU or something to ensure it. Stock clocks would be appreciated.
Also, run Vellamo benchmark too please. Linpack too lol. And NOT quadrant.
That said, Darkside is the best performer here so far.

2hvy4grvty said:
Also, run Vellamo benchmark too please. Linpack too lol. And NOT quadrant.
Click to expand...
Click to collapse
Sorry, that's not going to happen. While I'm still debating how I feel about stock vs a standard clock speeds, I can tell you that I'm not willing to try different benchmarking tools. I feel AnTuTu is modern enough. And I agree with the quadrant comment.

Why not? What else tests browsing performance as thoroughly?
Can we at least agree that despite all the little games and gadgets this phone runs, almost everyone primarily uses the phone to browse the Internet? Antutu does none of that.

Not to defend myself or anything but...
a) for the 10000th time benchmarks are useless
b) benchmarks are useless
c) you ran them after different clock speeds.
d) benchmarks are better for Kernel tests, not rom tests.
e) Australia.

For clarity, the goal wasn't to try to get the highest score. It was to try to put all ROM's (and their included kernels) at the same baseline and see how they perform relative to one another. Not trying to put one over another.
Finding the right ROM can sometimes be a daunting task. Many offer roughly the same features, but just have different appearance. When people don't care about how the ROM looks (with all the different launchers and mods that can be changed after the fact), knowing how it performs out of the box may be the deciding factor.
I know clock speeds are different. Some ROMs opt for lower clock speeds to get better better life. So I thought it would be good to know how that effects its performance.
With that said, would this be useful? Would anyone appreciate seeing how one ROM performs relative to the others?

But what I don't get is you claim that the goal of the thread was to put all ROMs at the same baseline. That didn't happen. Just pointing that out.
matthewdavis said:
For clarity, the goal wasn't to try to get the highest score. It was to try to put all ROM's (and their included kernels) at the same baseline and see how they perform relative to one another. Not trying to put one over another.
Finding the right ROM can sometimes be a daunting task. Many offer roughly the same features, but just have different appearance. When people don't care about how the ROM looks (with all the different launchers and mods that can be changed after the fact), knowing how it performs out of the box may be the deciding factor.
I know clock speeds are different. Some ROMs opt for lower clock speeds to get better better life. So I thought it would be good to know how that effects its performance.
With that said, would this be useful? Would anyone appreciate seeing how one ROM performs relative to the others?
Click to expand...
Click to collapse
Sent from my SGH-T989 using XDA App

I think he means baseline as in the performance you get from untouched roms as the dev intended them to be. This is probably more useful for people that are scared to mess around with overclocking apps..
G1ForFun said:
But what I don't get is you claim that the goal of the thread was to put all ROMs at the same baseline. That didn't happen. Just pointing that out.
Sent from my SGH-T989 using XDA App
Click to expand...
Click to collapse

mixedguy said:
I think he means baseline as in the performance you get from untouched roms as the dev intended them to be. This is probably more useful for people that are scared to mess around with overclocking apps..
Click to expand...
Click to collapse
That is correct. The baseline as defined by the authors of the ROMs. If a ROM author decides to underclock it out of the box, I'm not going to modify that.

[HOWTO] Detecting the LP Core + Disabling the other 4 (maybe)!

Nothing in this post should be considered gospel; These are simply conclusions I've drawn from testing. Everything written here could be (and probably is) very, very wrong!
Click to expand...
Click to collapse
YOU MUST BE ROOTED TO ATTEMPT ANYTHING IN THIS THREAD
I don't take full credit for this, someone posted information about the file I listen to but I can't for the life of me find the post. If someone locates it let me know so I can give credit where it's due!
Detecting the Companion Core
The 5th Core (Companion Core/Low Power - LP - Core) is invisible to the android system and cannot be controlled at the moment. However, the kernel can see the 5th core and can detect whether it's running or not.
This is a simple demonstration of this, nothing particularly interesting or useful yet!
Results with screen ON
{
"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"
}
Results with screen OFF
So yeah, this is why we get such great screen-off battery life!
I know pretty much everything written here is already known I just thought it would be nice to see some proof!
For those of you that want to try this out yourself, make a service that runs in the background that reads from: "/sys/kernel/cluster/active". It will be either "G" or "LP" (G meaning it's running from the normal cores, LP is the Low Power core.)
Interestingly, while the screen is on but nothing much is happening on the device it will constantly switch between G and LP (like in the first screenshot) but when doing CPU intensive tasks it runs as G constantly. When the screen is off, it never switches to G which is giving us the insane screen-off battery life.
Disabling 4 Cores and just running the LP Core!
NOTE: This may not be the case. The phone kinda gets a bit screwy and reports different things. I'm pretty sure this is just the throttle that happens when the CPU hits 90°c but I'm not sure (it's hard to make it hit 90°c!). If anyone knows what else it does please let me know!
WARNING: This can, and most probably will, crash your phone after a while. When running CM9 I even had to reflash the ROM (This hasn't happened on Android Revolution yet). Basically PLEASE make a backup and don't blame me if you mess anything up!
This is a strange one. If you navigate to "/sys/kernel/debug/cpu-tegra" there's a file called 'Throttling'. Changing this file to '1' makes the system throttle CPU usage drastically.
Interestingly though, it also makes the "sys/kernel/cluster/active" file report nothing but 'LP' even when doing tasks, indicating that it's running solely off the LP Core. However, I'm using TegraStats to monitor what's active and what isn't and when throttling is enabled it reports that 'cpu0' is also active, contrary to what the "sys/kernel/cluster/active" file says.
If you apply this mod you will instantly see why I think it's running soley off the LP Core: It makes the phone almost unusable and I wouldn't recommend doing it for any great length of time!
Another wierd thing is that TegraStats also reports that CPU1 activates at random points while the phone stays as slow as ever.
Of course, CPU1 could indeed be active and just throttled to a low clock speed (102mhz is pretty much the average when doing anything. It can go as high as 304 though) but it still doesn't explain the LP Core results.
So, if this is running off only the companion core, what does it mean? Well, you could ultimately gain massive battery improvements. The only problem is, the mod renders the phone unusable as it stands. The throttle would probably need to be modified from kernel source to not be quite as aggressive or something, so not useful at the moment but it has potential
I hope you found this interesting!

Reserved for future info!

Nothing to add, sorry, but very cool.

how to disable lp core.
i need to disable it completly. it has ****ty performance
or maybe you know how to change governor on that core or something with him

kragnegrozor said:
how to disable lp core.
i need to disable it completly. it has ****ty performance
or maybe you know how to change governor on that core or something with him
Click to expand...
Click to collapse
it's not good to disable the core because after that you wont have good standy battery life.....and what do you mean with it has bad performance?

man, newest update in youtube APP uses only 1 core, and this core is this sh****ty fifth one. in 720p films there are stuttering and other sh** sometimes. If you demand something powerful from phone at this time, other 4 cores are ON and video is playing smoothly.

Lolwut? The companion core should only be used once your screen is off or goes into deep sleep. Not for the main system to use or for playing YouTube videos.
Sent from my HTC One X using Tapatalk 2

Well done on bumping a 7 month old thread that obviously isn't relevant any more, guys...

so can somebody tell me how to disable it or change governor? it is for research

kragnegrozor said:
so can somebody tell me how to disable it or change governor? it is for research
Click to expand...
Click to collapse
Try System Tuner or any similar app.

kragnegrozor said:
so can somebody tell me how to disable it or change governor? it is for research
Click to expand...
Click to collapse
neuTrue said:
Try System Tuner or any similar app.
Click to expand...
Click to collapse
no app can disable it until a kernel dev can something do for it ...also nothing else can disable it ...try to search for example for youtube a fix that it will work better ...because the 5th core doesn't work in youtube or other ...youtube is stuttering because of other things not the 5th core

Do people know that the lp core has the exact same performance as the other cores?
Sent from my faster than SGS3 HOX.

XxVcVxX said:
Do people know that the lp core has the exact same performance as the other cores?
Sent from my faster than SGS3 HOX.
Click to expand...
Click to collapse
No it doesn't.
Sent from my HTC One X using Tapatalk 2

Learn about the topology of tegra3 unit. Different core technology, Different performance.
Wysyłane z mojego HTC One X za pomocą Tapatalk 2

XxVcVxX said:
Do people know that the lp core has the exact same performance as the other cores?
Sent from my faster than SGS3 HOX.
Click to expand...
Click to collapse
it has only 500 mhz and it will be used for apps which wont take much or if the screen is off/deep sleep...it wont be used in apps like youtube and other ...
like messaging/sms app or e-mail or so

One-X-master said:
it has only 500 mhz and it will be used for apps which wont take much or if the screen is off/deep sleep...it wont be used in apps like youtube and other ...
like messaging/sms app or e-mail or so
Click to expand...
Click to collapse
It has the exact same performance.
It's still ARM Cortex A9, only manufactured with low power silicon. It doesn't scale well into high clock speeds so thus its limited to 500 MHz, but it still has the same performance at same frequency.
Sent from my faster than SGS3 HOX.

XxVcVxX said:
It has the exact same performance.
It's still ARM Cortex A9, only manufactured with low power silicon. It doesn't scale well into high clock speeds so thus its limited to 500 MHz, but it still has the same performance at same frequency.
Sent from my faster than SGS3 HOX.
Click to expand...
Click to collapse
yes but it it isn't working while apps like youtube I have no problems when i use youtube he also had to write which kernel he use because some of the custom kernels or the stock don't know what it was have problems in some apps also figuered that out i had also with a kernel problems in youtube in other no problems...but don't know which because i don't have the phone here because of the f*cking repair center...

i am using kernels
,doing some research about performance.
Just some observations for now
i am need to know how to disable it.
Got some knowledge from many other phones\kernels\roms
But here is tegra3 topology

kragnegrozor said:
i am using kernels
,doing some research about performance.
Just some observations for now
i am need to know how to disable it.
Got some knowledge from many other phones\kernels\roms
But here is tegra3 topology
Click to expand...
Click to collapse
you know that EVERY android build has kernels because without they can't work? i meant with stock kernel that's the kernel in the stock RUU/ rom and custom kernels were edited by devs like faux etc...you CAN'T disable it try to ask a kernel dev you can see them in original development thread but like i know it's impossible because the 5th is for android unvisible ^^

i am gonna think about that.
i am seen somewhere OCed 5th core. Gonna find where and ask in that kernel thread.
thanks

source of lag due to zram?

I've been experiencing random lagging with my nexus 5x and I just noticed that the stock nexus 5x kernel implements zram.
Here's screenshot:
{
"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"
}
For those that don't know, zram is basically a compressed in RAM swap space.
Everytime the kernel saves to and from zram, it has to compress/decompress the contents and that could be the source of some of the lag people are seeing.
https://en.wikipedia.org/wiki/Zram
I plan on rooting my phone as soon as I get my usb c to usb a cable and disable zram and see if that helps with the random lags that i've been experiencing.

Very interesting theory....I am experiencing lag too with my 5X. It is very frustrating to compare with my two year old Nexus 5 which performs lightening fast and noticeably smoother.
Keep us posted when you root and disable zram :good:

The fact that google added the swap space seems to confirm that 2gb isn't enough for these 64bit SOCs. They shouldn't have skimped.

dwang said:
The fact that google added the swap space seems to confirm that 2gb isn't enough for these 64bit SOCs. They shouldn't have skimped.
Click to expand...
Click to collapse
If there is zram enabled in the 6P, then Google must have seen other problems. This is interesting though, thanks for informing us

0.0 said:
If there is zram enabled in the 6P, then Google must have seen other problems. This is interesting though, thanks for informing us
Click to expand...
Click to collapse
Good point.... Can anyone confirm is zram is enabled in the 6P?

Hi
dwang said:
The fact that google added the swap space seems to confirm that 2gb isn't enough for these 64bit SOCs. They shouldn't have skimped.
Click to expand...
Click to collapse
That's not the reason. Swap space takes up memory! The reason it is added is because it protects the flash memory and improves performance. Rather than having the swap drive all on flash memory, it is compressed and kept in RAM (because 2Gig is plenty), this improves performance because any pages swapped out are swapped to DRAM which is considerably faster (and uses less power), and saves lots of small writes to the flash memory, which stops it wearing out.
Google and LG haven't just thrown the Nexus 5X together without knowing what they are doing
Regards
Phil

seems like the encryption is the source of the bottleneck. the encryption with no easy way to disable. i cant wait till sources are released and i can get off android.

That makes no sense.

Hi
m4r0v3r said:
seems like the encryption is the source of the bottleneck. the encryption with no easy way to disable. i cant wait till sources are released and i can get off android.
Click to expand...
Click to collapse
I've disabled encryption (so I could root it) and the phone performs identical, besides I'm not seeing any lag or problems. Only user data is encrypted anyway, it's not the whole device that is encrypted, so any performance loss is negligible as it is only a tiny amount of data that is going through a decrypt/encrypt process.
The Nexus 5X is as fast for me as the graphical animations allow, if people are seeing lag have they tried a factory reset, leaving the phone to update all the apps without choosing to restore previous apps/data and trying the phone out? It could be an app problem/service that isn't compatible and eating CPU cycles.
Regards
Phil

I remember ZRam being used in the Jolla phone by some of the community, and in that instance it actually improved the performance.

My experience was the opposite. Back in the g2 days there were roms that enabled zram and it was smooth at first but then it got laggy after a few days.
I'm coming from the nexus 6, and the nexus 6 is smoother than the nexus 5x. This is in terms of input latency, scrolling, and load times.
It may very well that I have a defective phone, but a lot of other people have reported these issues as well. If you aren't experiencing this, more power to you, but don't disregard other people's experience.

dwang said:
My experience was the opposite. Back in the g2 days there were roms that enabled zram and it was smooth at first but then it got laggy after a few days.
I'm coming from the nexus 6, and the nexus 6 is smoother than the nexus 5x. This is in terms of input latency, scrolling, and load times.
It may very well that I have a defective phone, but a lot of other people have reported these issues as well. If you aren't experiencing this, more power to you, but don't disregard other people's experience.
Click to expand...
Click to collapse
Whatever the cause is, I've got this problem and I'm considering returning it as faulty and using my Nexus 5 for another year or two, because it's really bugging me now.

PhilipL said:
Hi
That's not the reason. Swap space takes up memory! The reason it is added is because it protects the flash memory and improves performance. Rather than having the swap drive all on flash memory, it is compressed and kept in RAM (because 2Gig is plenty), this improves performance because any pages swapped out are swapped to DRAM which is considerably faster (and uses less power), and saves lots of small writes to the flash memory, which stops it wearing out.
Google and LG haven't just thrown the Nexus 5X together without knowing what they are doing
Regards
Phil
Click to expand...
Click to collapse
Why there should be any swap space in the first place? Zram does indeed improve multitasking on systems with limited amounts of ram but there should be no reason to put it on an expensive device with "plenty of ram". Zram comes at a cost, cpu power and usable ram that is. If you do multi task a lot it's going to use more resources than necessary to keep up. If you are using an app that requires lots of ram it will slow your device down and cause a lot of heat because it is compressing/decompressing active pages. This is the reason most reviewers report that 5x is slower than the G4 (that has 1 more gb of ram) even though they have the same SoC.
It's the same Nexus 9 story all over again isn't it? Once again they used the absurd amount of 500+ MB which will cause the Low Memory Killer to kill everything except for the foreground app because there is no free ram.
But why did they do that? The 64bit android builds consume much much more ram than 32bit. If you disable zram on nexus 9 you will notice that it heats up much less and that it is faster with ram hungry applications but apps start dying even faster. Chrome will crash even with 1 tab open if the website is heavy. Launcher redraws will be everywhere. So we go from borderline usable to absolutely terrible.
I think they should have used 32bit builds on their 2gb devices. They would be faster and much more usable. There are no immediate benefits with 64bit anyway...

Hi
kdoul said:
Why there should be any swap space in the first place?...
Click to expand...
Click to collapse
Because it is breathing space should memory run short for whatever reason. We don't know the ZRAM is even being used, we just know it is there. There is a lot of complicated stuff going on with the Linux kernel and memory management, but typically in order to not waste memory, memory gets filled with all sorts of cached data until it is fall. Think of memory like an infinity swimming pool where the water goes right to the top edge, it doesn't matter how small or large that pool is, the whole point is to fill it to the brim. If you look over the edge of an infinity pool it has a catch channel and drain, because as soon as you use it, some water gets pushed out because it is fall. Memory is the same, because the idea is to use it all, there is always something getting pushed out, and rather than discard that completely, it gets swapped out, so when there is space again, it can be brought back in as required. The ZRAM is a friendly way of doing that to avoid lots of writes to flash memory, and because it is compressed, we might lose 500Meg to ZRAM, but the compression means it holds more than is lost, so we get a net gain.
Generally as I understand it, it's main purpose is to stop a constant trickle of writes to flash memory, which otherwise is wear on the flash memory and uses more power. It's job is a buffer for all the ebs and flows of a complex memory management system.
All major operating systems do this sort of thing, regardless of the amount of RAM available, because if you want to fill something to the brim, you've got have a way of dealing with an overflow in the most efficient manner.
You are correct in that 64bit is really a complete waste in smart phones and uses more memory, we have it because it's marketing, however the move to 64bit also brings some new instructions that aren't in the 32bit architecture in ARM, so there are improvements to be had as a consequence, but we don't need 64bit in smartphones really.
Regards
Phil

zram is being used by the phone. You can just do a "free" from the terminal and you can see that the zram swap space is being used. I have zram disabled on my 5x, or I would paste the output here, but its very easy to do even without root.
Also, there is no swap space on the flash enabled regardless of whether zram is enabled or not, so there is no risk that the flash memory will wear out.
If you disable zram, it doesn't mean that a swap space is automatically enabled on the flash.
PhilipL said:
Hi
Because it is breathing space should memory run short for whatever reason. We don't know the ZRAM is even being used, we just know it is there. There is a lot of complicated stuff going on with the Linux kernel and memory management, but typically in order to not waste memory, memory gets filled with all sorts of cached data until it is fall. Think of memory like an infinity swimming pool where the water goes right to the top edge, it doesn't matter how small or large that pool is, the whole point is to fill it to the brim. If you look over the edge of an infinity pool it has a catch channel and drain, because as soon as you use it, some water gets pushed out because it is fall. Memory is the same, because the idea is to use it all, there is always something getting pushed out, and rather than discard that completely, it gets swapped out, so when there is space again, it can be brought back in as required. The ZRAM is a friendly way of doing that to avoid lots of writes to flash memory, and because it is compressed, we might lose 500Meg to ZRAM, but the compression means it holds more than is lost, so we get a net gain.
Generally as I understand it, it's main purpose is to stop a constant trickle of writes to flash memory, which otherwise is wear on the flash memory and uses more power. It's job is a buffer for all the ebs and flows of a complex memory management system.
All major operating systems do this sort of thing, regardless of the amount of RAM available, because if you want to fill something to the brim, you've got have a way of dealing with an overflow in the most efficient manner.
You are correct in that 64bit is really a complete waste in smart phones and uses more memory, we have it because it's marketing, however the move to 64bit also brings some new instructions that aren't in the 32bit architecture in ARM, so there are improvements to be had as a consequence, but we don't need 64bit in smartphones really.
Regards
Phil
Click to expand...
Click to collapse

PhilipL said:
~text~
Click to expand...
Click to collapse
Memory management on your desktop or laptop computer is vastly different than that of android. On a desktop you try not to kill anything, because if you do something might break. If you run out of ram it is preferable to use some sort of swap space (pagefile in windows) in your hard drive to temporarily store inactive bits of memory until you need to use them later. It is slow but it gets the job done.
On android devices, storage is (maybe used to be) slow to read and write on so you want to use it the least possible. The solution is to maximize ram usage and fill it all the way at all times. If you have no apps open and lots of unused ram android will start caching your most used apps in the unused space to reduce loading times. If you do have apps opened they will stay frozen in the background until you access them again. But what happens when the unused ram goes low? Ask a special program to go on a killing spree. The idea is that you care about what you see and everything past that is not critical. Frozen apps die first, then some background services, then the foreground services and at last, the foreground app.
Then there comes the zram. The idea is that you cut a part of your physical ram and make it work like the swap partition on your hard dive, but you compress the data first before you put them there. Since ram is much faster than your hard drive zram, if configured correctly, can save you from swap slowdowns when using more than 100% of your physical ram.
Zram is problematic on android. How do you decide what to swap and what to kill when you are low on memory?
Battery and processing power are also limited. Heat when constantly using the cpu for compression is an issue that can cause slowdowns because of thermal throttling. On this scenario you are taking a huge piece for that purpose (more than 1/4) of the available physical ram when you should be taking less than half of that.
The only acceptable place for zram on android is on budget phones/tablets that simply can't afford to have more than 1gb of ram and you want them to run memory hungry applications without crashing. But then again, they do slow down when it is used. A lot.
When you have an expensive (flagship) device and you see that it can't run properly with the amount of ram you have already supplied the only way to ensure it will work optimally is to add more freaking ram. Not enable zram or any other software feature to try and bypass that. Google was stupid with the Nexus 9 and once again they are being stupid with the 5X. Cheaper or equally priced devices have tons of ram nowadays so I just can't see a reason other than Google wanting to either save peanuts from using less or promote their higher end model because it feels faster???
I can accept bad cameras, ugly design, small batteries, ****ty audio or even mediocre processing power. What I can't accept is (purposely?) crippled devices that they are broken out of the box in a way that you can't possibly fix.
Small edit: I believe that you can use the new crypto instructions from the ARMv8 instruction set and still run android in Aarch32 (32 bit) mode. In that case everyone is happy except for the marketing team of course.

So I've been running with zram disabled (in addition to disabling encryption) for about a day now, and I've not noticed any problems. The random lag I've been experiencing before are now largely gone and the phone is just as fast and smooth as my n5 and n6.
When i first got the n5x and experiencing the random lag, I was seriously considering returning it, but now that I've disabled zram and decrypted the phone, I'm fairly happy now. Though there are still issues related to charging and battery life.

If you don't use live wallpapers, facebook chatheads and keep the background junk in check it should work just fine. If you have problems with chrome killing tabs too often, try firefox.
Custom kernels should help with the rest.

okay im unencrypted and rooted, how can i disable zram? I've looked everywhere and cant find an answer. thanks!

Hey Guys I am unlocked but not rooted as i want to keep using android pay and keep my phone encrypted. Some here said there is a way to disable zram without root...how would one do that?
MY girls nexus 5x is lagging pretty hard and just an FYI my Nexus 6P also has ZRAM enabled. i like to turn that off as well. No need for it with 3gb ram!