So I saw this over on the s3 forums and it gave me an idea
If everyone who sees this could post the kernel setup (overclocking, schedulers, or other tweaks) that has given either a) the best performance or b) the best battery life, it would be awesome and really help noobs like me who have never overclocked :laugh:
I know that every phone is different and blah blah blah, but I'm hoping that this would at least give us kernel tweaking noobs someplace to start besides just changing random settings and hoping for the best.
Thank you all in advance, all contributions are welcome
When posting, please state if you are addressing performance or battery life tweaks, and what Android version those were applied on just to make things a bit easier to follow
:good: :good:
Thanks!
TWEAKS
Performance
1.5 max, 255 min Interactive noop - Contributed by @thronnos
1.728 max, 184 min Performance noop - Contributed by @klund29078
Balanced
1.2 Max, 245 min Interactive noop - Contributed by @Chad the Pathfinder
I run 1.5/255 Interactive Noop
Shifted from planet Namek
---------- Post added at 10:21 PM ---------- Previous post was at 10:21 PM ----------
But I also am running a couple PmR and Seeder tweaks
Shifted from planet Namek
Like I said, please indicate if thats for performance, or battery life, just helps to sort things out
Sent from my PG06100
CNexus said:
Like I said, please indicate if thats for performance, or battery life, just helps to sort things out
Sent from my PG06100
Click to expand...
Click to collapse
Oh its been the best all around performance I could find. Only thing that ran generally better was an old sparksco kernel. And with wifi the phone can last days
Shifted from planet Namek
Cmon guys, theres gotta be more people who have good settings.
Sent from my PG06100
I'm by no means an expert on the subject but I run:
1.728/184 performance noop.
Honestly, I'd like to know the meanings of all the governors and I/o schedulers to have a better understanding of it. Mine is set for performance, obviously, but the battery life is not that bad. I always keep a charger with me though because I use my phone heavily everyday.
CNexus said:
Cmon guys, theres gotta be more people who have good settings.
Sent from my PG06100
Click to expand...
Click to collapse
I got tons of stuff, just most of it's not very organized... I'll try to post some tweaks and such up here later when I get off work. I wish there was a tool to organize my computer and SD card for me... But it probably wouldn't do it the way I want
klund29078 said:
I'm by no means an expert on the subject but I run:
1.728/184 performance noop.
Honestly, I'd like to know the meanings of all the governors and I/o schedulers to have a better understanding of it. Mine is set for performance, obviously, but the battery life is not that bad. I always keep a charger with me though because I use my phone heavily everyday.
Click to expand...
Click to collapse
I got you
http://forum.xda-developers.com/showthread.php?t=1941195
Click the "Android CPU governors explained" link
Sent from my PG06100
I've tried several combonations and the one I like most on my Shift is 1.2 max 245 min Interactive, noop. That seems to give me the best ballance between battery life and performance. My favorite kernels for sense ROMs have been Strapped's Supreme Sense kernel and Dodge's v1.6. Others are good too, these are just my favorites.
CNexus said:
I got you
http://forum.xda-developers.com/showthread.php?t=1941195
Click the "Android CPU governors explained" link
Sent from my PG06100
Click to expand...
Click to collapse
you the man! thanks for the link.
Here's some good build.prop tweaks:
ro.max.fling_velocity=15000
ro.min.fling_velocity=10000
ro.HOME_APP_ADJ=1
debug.sf.hw=1
debug.performance.tuning=1
debug.composition.type=gpu
debug.kill_allocating_task=1
debug.egl.hw=1
video.accelerate.hw=1
persist.sys.use_dithering=1
persist.sys.purgeable_assets=1
persist.sys.shutdown.mode=hibernate
pm.sleep_mode=1
ro.ril.disable.power.collapse=0
ro.ril.hep=1
ril.ecclist=911,#911,*911
media.stagefright.enable-player=true
media.stagefright.enable-meta=true
media.stagefright.enable-scan=false
media.stagefright.enable-http=true
ro.mot.eri.losalert.delay=1000
ro.lge.proximity.delay=25
mot.proximity.delay=25
dalvik.vm.execution-mode=int:jit
dalvik.vm.dexopt-data-only=1
ro.compcache.default=0
net.tcp.buffersize.wimax=4096,221184,524288,4096,16384,110208
net.tcp.buffersize.gprs=4094,87380,256960,4096,16384,110208
net.tcp.buffersize.wifi=4096,87380,256960,4096,16384,256960
You can copy and paste them into build.prop with a root browser, save, and reboot
---------- Post added at 06:32 PM ---------- Previous post was at 05:50 PM ----------
Here's some more tweaks you can put in sysctl.conf (or make it if you don't have it):
net.ipv4.tcp_ecn=0
net.ipv4.route.flush = 1
net.ipv4.tcp_rfc1337 = 1
net.ipv4.ip_no_pmtu_disc = 0
net.ipv4.tcp_sack = 1
net.ipv4.tcp_fack = 1
net.ipv4.tcp_window_scaling = 1
net.ipv4.tcp_timestamps = 1
net.ipv4.tcp_rmem = 4096 39000 187000
net.ipv4.tcp_wmem = 4096 39000 187000
net.ipv4.tcp_mem = 187000 187000 187000
net.ipv4.tcp_no_metrics_save = 1
net.ipv4.tcp_moderate_rcvbuf = 1
net.ipv4.conf.all.accept_source_route = 0
net.ipv4.conf.all.accept_redirects = 0
net.ipv4.conf.default.accept_redirects = 0
net.ipv4.conf.all.secure_redirects = 0
net.ipv4.conf.default.secure_redirects = 0
kernel.sched_child_runs_first=0
kernel.sched_rt_period_us=1000000
kernel.sched_rt_runtime_us=950000
kernel.panic=30
kernel.core_uses_pid=0
kernel.core_pattern=core
kernel.core_pipe_limit=0
kernel.tainted=512
kernel.real-root-dev=0
kernel.print-fatal-signals=0
kernel.ctrl-alt-del=0
kernel.ftrace_enabled=1
kernel.ftrace_dump_on_oops=0
kernel.threads-max=10000
kernel.random.poolsize=4096
kernel.random.read_wakeup_threshold=512
kernel.random.write_wakeup_threshold=256
kernel.overflowuid=65534
kernel.overflowgid=65534
kernel.pid_max=32768
kernel.panic_on_oops=1
kernel.dmesg_restrict=0
kernel.kptr_restrict=0
kernel.ngroups_max=65536
kernel.watchdog=1
kernel.watchdog_thresh=10
kernel.softlockup_panic=1
kernel.nmi_watchdog=1
kernel.randomize_va_space=1
kernel.hung_task_panic=0
kernel.hung_task_check_count=32768
kernel.hung_task_timeout_secs=30
kernel.hung_task_warnings=10
kernel.max_lock_depth=0
kernel.keys.maxkeys=200
kernel.keys.maxbytes=20000
kernel.keys.root_maxkeys=200
kernel.keys.root_maxbytes=20000
kernel.keys.gc_delay=300
kernel.perf_event_paranoid=1
kernel.perf_event_mlock_kb=516
kernel.perf_event_max_sample_rate=150000
kernel.blk_iopoll=1
kernel.hostname=localhost
kernel.domainname=localdomain
kernel.shmmax=33554432
kernel.shmall=2097152
kernel.shmmni=4096
kernel.msgmax=8192
kernel.msgmni=1041
kernel.msgmnb=16384
kernel.auto_msgmni=1
vm.min_free_kbytes=4096
vm.panic_on_oom=0
vm.oom_kill_allocating_task=0
vm.vfs_cache_pressure=20
vm.oom_dump_tasks=1
vm.drop_caches=3
vm.max_map_count=65530
vm.block_dump=1
vm.stat_interval=1
vm.mmap_min_addr=204
vm.highmem_is_dirtyable=1
vm.scan_unevictable_pages=0
---------- Post added at 06:59 PM ---------- Previous post was at 06:32 PM ----------
Or here is the sysctl.conf file with the tweaks listed above. Place in system/etc with a root browser, fix permissions, and reboot
http://db.tt/LTFf4Vdi
Oh and the sysctl.conf file is meant for 4.1+ AOSP, make a nand first, I can't guarantee anything under that... Just make a nand. I don't see why it wouldn't work on lower versions but I don't know, I only run JB now.
bilgerryan said:
Oh and the sysctl.conf file is meant for 4.1+ AOSP, make a nand first, I can't guarantee anything under that... Just make a nand. I don't see why it wouldn't work on lower versions but I don't know, I only run JB now.
Click to expand...
Click to collapse
To support what bilgerryan said, I tried the sysctl.cont file on sense gb before and it didn't help.
Chad The Pathfinder said:
To support what bilgerryan said, I tried the sysctl.cont file on sense gb before and it didn't help.
Click to expand...
Click to collapse
Yeah just be careful with using it with Sense, I've seen it cause boot loops on TouchWiz (people never listen... Or is it that they just didn't read the instructions?). Either way, it's meant for JB AOSP and probably won't work on anything below ICS.
The build.prop tweaks work on all ROMs tho.
bilgerryan said:
Yeah just be careful with using it with Sense, I've seen it cause boot loops on TouchWiz (people never listen... Or is it that they just didn't read the instructions?). Either way, it's meant for JB AOSP and probably won't work on anything below ICS.
The build.prop tweaks work on all ROMs tho.
Click to expand...
Click to collapse
It worked wonders for me on GB
CNexus said:
It worked wonders for me on GB
Click to expand...
Click to collapse
The sysctl.conf or the build.prop tweaks? I'm talking about the specific sysctl.conf that I linked to above Just for clarification. You can use a sysctl.conf file on any ROM, just saying these tweaks I did for JB.
And as far as overclocking on the Shift I use the same settings as Chad except I use Smartass (or v2 if available) and sio or deadline for I/O
Related
I absolutely take no creative credit in making this kernel as all I have done is used the lower voltage code on Ivan Dimkovic's github and build based on the LEO kernel tree. The real credit goes to them.
Please check "Save battery while running ANY Android build on your HD2" for how to tweak the system with setcpu.
In this kernel, I have enabled the follows:
Conservative CPUfreq governor
Conservative governor is the default CPUfreq governor
Voltage goes down to 1000 mv (zImage_1000mv) or 925mv (zImage_925mv) or 1050mv (default voltage)
Overclocking to 1.113 GHz max.
Remember, I hold NO responsibility for anything that goes wrong. I'm simply providing the kernel, whatever happens with it is on your shoulders not mine.
If you do not like the overclocked feature, you can use SetCPU to disable it.
minor tweaking to drivers/cpufreq/cpufreq_ondemand.c
#define DEF_FREQUENCY_DOWN_DIFFERENTIAL (30)
#define MICRO_FREQUENCY_UP_THRESHOLD (65)
#define MICRO_FREQUENCY_MIN_SAMPLE_RATE (9500)
minor tweaking to drivers/cpufreq/cpufreq_conservative.c
#define DEF_FREQUENCY_UP_THRESHOLD (65)
#define DEF_FREQUENCY_DOWN_THRESHOLD (30)
To use it, overwrite the existing zImage and push the bcm4329.ko to the /system/lib/modules:
Code:
adb remount
adb push bcm4329.ko /system/lib/modules/bcm4329.ko
for detailed instruction, please check[Discussion]Linux and Android on the HD2 *READ POST4 BEFORE ASKING QUESTIONS
The main purpose of the AUFS is to provide an overlay to a read only file system (usually a compressed filesystem, such as squashfs or cramfs). Since the ext2 images are writable, and we are not using a compressed filesystem, what is the point of using AUFS? The real issue is this, currently the HD2 is running on the sdcard, and reading from sdcard is slow. If the file system is compressed, you'll read less and decompress it, and decompression is pretty fast. Hopefully, the decompression overhead is smaller than the read performance gain from a compressed file system. It could mean that overall compressed file system is faster than ext2 (i hope).
Change Log
version 0.4:
kernel based on LEO tree (HEAD version 827fa85),
add aufs support
huanyu said:
I absolutely take no creative credit in making this kernel as all I have done is used the lower voltage code on Ivan Dimkovic's github and build based on Cotulla kernel tree. The real credit goes to them.
In this kernel, I have enabled the follows:
Conservative CPUfreq governor
Conservative governor is the default CPUfreq governor
Voltage goes down to 0.925 volts
Overclocking to 1.1 GHz max.
You can use SetCPU to configure the overclocking settings
minor tweaking to drivers/cpufreq/cpufreq_ondemand.c
#define MICRO_FREQUENCY_UP_THRESHOLD (65)
#define MICRO_FREQUENCY_MIN_SAMPLE_RATE (9500)
minor tweaking to drivers/cpufreq/cpufreq_conservative.c
#define DEF_FREQUENCY_UP_THRESHOLD (65)
#define DEF_FREQUENCY_DOWN_THRESHOLD (30)
To use it, overwrite the existing zImage and push the bcm4329.ko to the /system/lib/modules:
Code:
adb remount
adb push bcm4329.ko /system/lib/modules/bcm4329.ko
Click to expand...
Click to collapse
interesting... did you include manusfreedom's 384mb RAM patch by any chance?
pongster said:
interesting... did you include manusfreedom's 384mb RAM patch by any chance?
Click to expand...
Click to collapse
lol.. only 1 way to find out..
Ok nice, I guess. What does it bring me?
Better battery performance probably. Do you have any experience with it? Does it slow down the overall performance?
Im willing to test it, but i need to know what to look for...
Sent from my HTC HD2 using XDA App
How about a detailed step by step NOOB guide to use this. Where to put the downloaded file, what to use to overwrite, etc. may sound like stupid questions but....
Thanks
Is this bcm4329.ko incluying the wifi module for ZImage.384Mo ?
Thxs.
huanyu said:
I absolutely take no creative credit in making this kernel as all I have done is used the lower voltage code on Ivan Dimkovic's github and build based on Cotulla kernel tree. The real credit goes to them.
In this kernel, I have enabled the follows:
Conservative CPUfreq governor
Conservative governor is the default CPUfreq governor
Voltage goes down to 0.925 volts
Overclocking to 1.1 GHz max.
You can use SetCPU to configure the overclocking settings
minor tweaking to drivers/cpufreq/cpufreq_ondemand.c
#define MICRO_FREQUENCY_UP_THRESHOLD (65)
#define MICRO_FREQUENCY_MIN_SAMPLE_RATE (9500)
minor tweaking to drivers/cpufreq/cpufreq_conservative.c
#define DEF_FREQUENCY_UP_THRESHOLD (65)
#define DEF_FREQUENCY_DOWN_THRESHOLD (30)
To use it, overwrite the existing zImage and push the bcm4329.ko to the /system/lib/modules:
Code:
adb remount
adb push bcm4329.ko /system/lib/modules/bcm4329.ko
Click to expand...
Click to collapse
Had try it but call, data do not works
Overclocking works however linpack scores arent any different, the undervolting works too (voltage readings on bootup are lower). Thanks for the kernel
p.s. you should try adding this to the 384mb ram image here: http://htcleo.noobhands.net/manusfreedom/
JJbdoggg said:
Overclocking works however linpack scores arent any different, the undervolting works too (voltage readings on bootup are lower). Thanks for the kernel
p.s. you should try adding this to the 384mb ram image here: http://htcleo.noobhands.net/manusfreedom/
Click to expand...
Click to collapse
thnks for the share
JJbdoggg said:
Overclocking works however linpack scores arent any different, the undervolting works too (voltage readings on bootup are lower). Thanks for the kernel
p.s. you should try adding this to the 384mb ram image here: http://htcleo.noobhands.net/manusfreedom/
Click to expand...
Click to collapse
The 1.113GHz has about 30 linpack score ( pretty consistent ) and the default speed (998MHz) is about 28. Make sure you switch to performance mode when running linpack.
WOW!! THIS IS AWESOME!! Anyway to combine this with the extra ram zImage?! For those not getting the overclock to work, you must redo autodetect speeds.
huanyu said:
... Make sure you switch to performance mode when running linpack.
Click to expand...
Click to collapse
Performance mode????? Where is this thing?
OrionBG said:
Performance mode????? Where is this thing?
Click to expand...
Click to collapse
You need SetCPU app to set overclock and other profiles. It is free for xda membs and a paid app in the market. My favorite app so far.
Link-->http://forum.xda-developers.com/showthread.php?t=734886
reocej said:
You need SetCPU app to set overclock and other profiles. It is free for xda membs and a paid app in the market. My favorite app so far.
Link-->http://forum.xda-developers.com/showthread.php?t=734886
Click to expand...
Click to collapse
Oh! I though that there is some setting in linpack!
BTW Can somebody tell me why the HD2 without OC scores around 27 in Linpack and the Samsung Galaxy-S that have supposedly better CPU scores around 14 (under Froyo)???
I'd love to see these features go into a mainstream build. Especially the undervolting. It'd make it much easier for people like me..
JJbdoggg said:
p.s. you should try adding this to the 384mb ram image here: http://htcleo.noobhands.net/manusfreedom/
Click to expand...
Click to collapse
i will wait until the 384mb ram merge into the LEO kernel.
aceo07 said:
I'd love to see these features go into a mainstream build. Especially the undervolting. It'd make it much easier for people like me..
Click to expand...
Click to collapse
I have create a patch for the undervolt and overclock features. Other kernel developers/builders can just apply the patch to have the UV and OC features.
Someone please implement this with the increased ram kernel!
reocej said:
Someone please implement this with the increased ram kernel!
Click to expand...
Click to collapse
Its a kernel not a hack
@OP
Add [KERNEL] to your post title, it'd give it alot more views.
Updated my thread on battery saving and added a link and directions to using the conservative power setting enabled in your kernel. Check it out and let me know if i missed something. Maybe add it to the first post to save you some questions.
Thanks for this!
just for curiosity sake, whats the original voltage?
and how do i push the bcm module?
Any update on this kernel? Dying to see this merged with the fullram kernel!!!!
This discussion aims at getting our x10 the best ROM or kernel or any MOD of all TIME!!
Benchmarking is an easy way to test your phone performance.Either CPU,2D & 3D graphics or SD card and Data transfer.
you can do it simply by software like AnTuTu or Quadrant.
The Point in this discussion is to test your phone and write ur best score at all time.Then you tell us your phone status.and tell us what is -in your opinion is the cause of your high score,so that we can determine which are best ROMs,Kernels or CPU governors,etc.
Please notice that Benchmarking is A ROUGH indicator of your phone performance. it is not necessarily that your phone is slower if you have lower scores
PLEASE SHARE YOUR
1.Best score.
2.which software did u use? scores are very different.
3.ROM.
4.kernel.
5.CPU min/max.
6.CPU governor.
7.SD card info.
8.Opinions.
Thanks,
Omar.
---------- Post added at 01:55 AM ---------- Previous post was at 01:37 AM ----------
MINE are:
1.1624
2.Quadrant.
3.WB 5.0c ROM.
4.DooMKernel v4.0.
5.192/1152 Mhz.
6.Performance.
7.8GB sandisk,class 6.
8. i think WB tweaks has increased my scores specially in v 5.0.
also overclocking and performance governor-CPU @ 1152 all the test-.
I was just about to ask this myself
1. 38.45 (Avg of 5 tests)
2. Linpack Pro
3. WB 5.0c ROM.
4. WB stock with OC tweak
5. 245/1152 Mhz.
6. interactive
7. 32GB Sandisk, class 4.
8. None.
It would be great if devs could post the performance stats of their Rom with their releases
1.- 2592
2.- AnTuTu Benchmark
3.- CM7 ZMOD
4.- DooMKernel v5 with overclock
5.- 128 - 1113 mhz
6.- smartassv2
7.- stock 16gb class 2
8.- I think you can get a better score with a tweaked rom, a better sd card and another cpu scaling
Score: 1795 Quadrant Advanced
ROM: MIUI 1.9.16 by 28Spawn with SWAP included- no data restored.
26fps stable on graphics- 2D leaf
class 2 stock 8gb card
Kernel: DooMKernel 05 for unlocked bootloader min 128 max 1075 smartass v2
8. SWAP files incrase my performance greatly. Tweaks do nothing but drain battery and cause lag. IMHO.
2352
Quadrant standard
Spaarc oxygenated gingerscape
Stock kernel
Locked bootloader
245-1113
Basic z oc
Zdzihu smartass
Memory card class 4
6.8 mB/s read
11.9 mB/s write
sent from my Android powered smart phone
1. 1.288
2. Quadrant Std.
3. Wolf 6.0.0c
4. DooMKernel v04b
5. 1.152 / 128
6. Interactive
7. stock 8GB
8. None
KuSkUS_68 said:
1. 1.288
2. Quadrant Std.
3. Wolf 6.0.0c
4. DooMKernel v04b
5. 1.152 / 128
6. Interactive
7. stock 8GB
8. None
Click to expand...
Click to collapse
I think your benchmarking should have a higher score.is that the highest one?
Sent from my X10i using XDA App
Prodigy said:
Score: 1795 Quadrant Advanced
ROM: MIUI 1.9.16 by 28Spawn with SWAP included- no data restored.
26fps stable on graphics- 2D leaf
class 2 stock 8gb card
Kernel: DooMKernel 05 for unlocked bootloader min 128 max 1075 smartass v2
8. SWAP files incrase my performance greatly. Tweaks do nothing but drain battery and cause lag. IMHO.
Click to expand...
Click to collapse
Yes man..i only got those scores with miui.
1.1755
2.Quadrant advanced.
3.Miui 1.9.9. Erenz's
4.DoomKernel v3b.
5.192/1152.
6.performance.
7.8GB sandisk class 6.
8. i got very high I/O scores "1600".used to get only 600 with WB ROM v5c.
but with AnTuTu are very low,only 4.5 Mb/s read & 8 Mb/s write. don't know why!!
Sent from my X10i using XDA App
---------- Post added at 01:29 PM ---------- Previous post was at 01:18 PM ----------
spaarc said:
2352
Quadrant standard
Spaarc oxygenated gingerscape
Stock kernel
Locked bootloader
245-1113
Basic z oc
Zdzihu smartass
Memory card class 4
6.8 mB/s read
11.9 mB/s write
sent from my Android powered smart phone
Click to expand...
Click to collapse
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
romasalah said:
Yes man..i only got those scores with miui.
1.1755
2.Quadrant advanced.
3.Miui 1.9.9. Erenz's
4.DoomKernel v3b.
5.192/1152.
6.performance.
7.8GB sandisk class 6.
8. i got very high I/O scores "1600".used to get only 600 with WB ROM v5c.
but with AnTuTu are very low,only 4.5 Mb/s read & 8 Mb/s write. don't know why!!
Sent from my X10i using XDA App
---------- Post added at 01:29 PM ---------- Previous post was at 01:18 PM ----------
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
Click to expand...
Click to collapse
Yes I'm using x10 very well ...
Search for gingerscape in x10 development thread
sent from my Android powered smart phone
romasalah said:
I think your benchmarking should have a higher score.is that the highest one?
Click to expand...
Click to collapse
Actually that's what I thought comparing yours.
I have been trying other governors and the best score till now is 1378 with smartassv2 (tried only interactivex, interactive and smartassv2)
I'll keep on trying with other settings...
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
http://forum.xda-developers.com/showthread.php?t=1212012&highlight=smartass
Skumtomten said:
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
http://forum.xda-developers.com/showthread.php?t=1212012&highlight=smartass
Click to expand...
Click to collapse
Dude it's the rom Not smartass
https://picasaweb.google.com/100227745563198630514/DropBox#5651119841938055298
romasalah said:
Yes man..i only got those scores with miui.
1.1755
2.Quadrant advanced.
3.Miui 1.9.9. Erenz's
4.DoomKernel v3b.
5.192/1152.
6.performance.
7.8GB sandisk class 6.
8. i got very high I/O scores "1600".used to get only 600 with WB ROM v5c.
but with AnTuTu are very low,only 4.5 Mb/s read & 8 Mb/s write. don't know why!!
Sent from my X10i using XDA App
---------- Post added at 01:29 PM ---------- Previous post was at 01:18 PM ----------
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
Click to expand...
Click to collapse
he's obviously using the io hack which makes this score fake.. .
you could not get a legitimate score that high on an x10...
---------- Post added at 04:22 PM ---------- Previous post was at 04:21 PM ----------
spaarc said:
Dude it's the rom Not smartass
https://picasaweb.google.com/100227745563198630514/DropBox#5651119841938055298
Click to expand...
Click to collapse
so your rom uses a hack to give fake scores, they are trying to figure out the fastest rom objectively.
disable the hack and report back
scoobysnacks said:
he's obviously using the io hack which makes this score fake.. .
you could not get a legitimate score that high on an x10...
---------- Post added at 04:22 PM ---------- Previous post was at 04:21 PM ----------
so your rom uses a hack to give fake scores, they are trying to figure out the fastest rom objectively.
disable the hack and report back
Click to expand...
Click to collapse
1376
That's the score after disabling the tweak
scoobysnacks said:
he's obviously using the io hack which makes this score fake.. .
you could not get a legitimate score that high on an x10...
---------- Post added at 04:22 PM ---------- Previous post was at 04:21 PM ----------
so your rom uses a hack to give fake scores, they are trying to figure out the fastest rom objectively.
disable the hack and report back
Click to expand...
Click to collapse
does MIUI uses this hack?
I got 1804 with MIUI 1.9.16
@1152-smartass v2
romasalah said:
does MIUI uses this hack?
I got 1804 with MIUI 1.9.16
@1152-smartass v2
Click to expand...
Click to collapse
no idea. anything over 1500 or 1600 is generally pretty suspect though
Skumtomten said:
how on smartass ?
That's very high man.. are using XPERIA x10 i need a photo to make sure u aren't using SGS II
http://forum.xda-developers.com/showthread.php?t=1212012&highlight=smartass
Click to expand...
Click to collapse
yeah,thanks I know about smartass and i use it. But i always have a wrong belief that performance is the best governor for benchmarking.May be cause it's always operates the processor @ Max frequency
---------- Post added at 06:18 PM ---------- Previous post was at 06:15 PM ----------
scoobysnacks said:
no idea. anything over 1500 or 1600 is generally pretty suspect though
Click to expand...
Click to collapse
i guess MIUI may be using it...with AnTuTu my benchmarking is normal -from 2500 to 2700- but with Quadrant my I/O has increased from 600 to 1600 after MIUI.
...more of a test...
1.
a) 2262 (3087)
b) 1274 (at the third times), (1440)
c) 37 MFLOPS (over 39)
d) 60 sec. (51 sec.)
e) 30 fps (31.2)
2.
a) Antutu benchmark.
b) Quadrant
c) Linpack
d) SQLite
e) Neocore
3.XperiaTW 6.0c ROM.
4.DooMKernel v4.0.b
5.128/1113 Mhz.
6.SmartassV2.
7.16GB,class 6 Kingstone.
8. I state that the test results, the present condition (full of applications, more than 100, then the X10 is under stress), are lower than those obtained in the past without so many applications installed. I have met with many peaks FROYO (TW custom ROM) and are shown alongside those of the scores in brackets .. I tried several tests in the past who now do not use hardly any, this thread I have "suggested" to review the scores to currently, although the comparison is not reasonable
Sorry for my English, I used google translate
1. 1.004
2. Quadrant Std.
3. Wolf 6.0.0c
4. DooMKernel v04b
5. stock
6. Im a Noob so dont know Its all at standart^^
7. stock 8GB
8. None
Cecco1970 said:
1.
a) 2262 (3087)
b) 1274 (at the third times), (1440)
c) 37 MFLOPS (over 39)
d) 60 sec. (51 sec.)
e) 30 fps (31.2)
2.
a) Antutu benchmark.
b) Quadrant
c) Linpack
d) SQLite
e) Neocore
3.XperiaTW 6.0c ROM.
4.DooMKernel v4.0.b
5.128/1113 Mhz.
6.SmartassV2.
7.16GB,class 6 Kingstone.
8. I state that the test results, the present condition (full of applications, more than 100, then the X10 is under stress), are lower than those obtained in the past without so many applications installed. I have met with many peaks FROYO (TW custom ROM) and are shown alongside those of the scores in brackets .. I tried several tests in the past who now do not use hardly any, this thread I have "suggested" to review the scores to currently, although the comparison is not reasonable
Sorry for my English, I used google translate
Click to expand...
Click to collapse
thanks for doing all these tests..
a large number of app affects memory testing and I/O..but i think not too much it doesn't have great range to play with.
Also i noticed high peaks with AnTuTu reaching above 3000 only with FROYO.
and noticed high graphics scores..may be double GB. but i don't know why .
Can't state that froyo has higher graphics than GB -although i heared alittle of that - Coz i didn't see high benchmarks with Quadrant in froyo.
Over my time here in XDA I've built up a few studies on the Nexus S.
Here are the links. There are summaries in each thread in the first and second (and sometimes third) posts of the main findings, but I've done some very quick and minimal overviews here also.
How to build a kernel: this thread
-This post takes you through installing Ubuntu right through to managing git, in order to compile your own kernel, and flash it in CWM recovery!
-Includes the tools you need to add MTP capabilities to your kernel.
User eXperience Benchmarks: this thread
-This shows which ROM and kernel threads have the best user feedback. The feedback is processed by an external sentiment analysis API. It is only meant as a rough guide, something of a quick reference particularly if you are looking for a new ROM to try. As of June 2012 this is no longer maintained.
All ICS ROM Benchmarks: this thread
-Freely available benchmark programs were used to determine which ROMs had the best performance. Generally the difference between the best ones is very small.
Battery Drain Benchmarks: this thread
-A collection of tips for getting the best battery life, and dispelling some myths.
Kernel Features and Benchmarks: this thread
-My original study, the one that started it all, the kernel benchmarks! Provides a quick table to compare which kernels perform best in CPU/RAM/IO/Database/Graphics.
CPU Governors and I/O Schedulers: this thread
-A study done to find which schedulers perform best in combination with the most popular governors. My two favourites (ondemand and smartassV2) combine best with noop.
Power Saving Governors: this thread
-A meta study done combining the findings from the battery study and governor study to show which governors will save the most power. Ondemand comes out significantly better than conservative, which surprised me. SmartassV2 is the best compromise for high performance and efficiency.
Does SuperCharging work?: this thread
-This was just a short study to find out if this script is suitable for the Nexus S, but it turned out it had no effect, positive or negative.
Kernel Memory Allocators: this thread
-This is another short study for kernel developers to show that SLUB performs best.
====
This thread is just an index. Our new mod veyka agreed would be a good idea to summarise all my studies for addition into the sticky roundup.
====
EDIT: Feel free to suggest any new things that I might be able to test!
im out of thanks for today.. but, awesome work regardless. thank you
simms22 said:
im out of thanks for today.. but, awesome work regardless. thank you
Click to expand...
Click to collapse
Me too, but thank you too!
Edit: ah just use the app, unlimited thanks!
bedalus said:
Me too, but thank you too!
Edit: ah just use the app, unlimited thanks!
Click to expand...
Click to collapse
really? unlimited? i dont use the app though, just the browser.
Uh, well, I haven't tested the limit... but I never get denied!
All your work was great I battery drain thread was very helpful
Sent from my Nexus S
Thanks for all your hard work.
IF , one of my assassins had done their job ..........
.. Generated via Tapatalk ..
Oodie said:
IF , one of my assassins had done their job ..........
.. Generated via Tapatalk ..
Click to expand...
Click to collapse
When they finally come for me, I will benchmark their performance.
bedalus said:
When they finally come for me, I will benchmark their performance.
Click to expand...
Click to collapse
nice answer
And nice guide for your very instructive benchmarks bedalus
Even when browsing through the app you'd still get that annoying thanks limit.
This is a really good overview for new NS users. (I doubt there are ever some since mostly would go directly for G-Nex)
What's next?
That "Oodie's Assassins' Performance Benchmark" sounds interesting, though.
PS: @Oodie, you should have kanged your assassins before sending them to bedalus' compound. That way, your assassins' would have at least injured bedalus
UPDATE: Memory Allocators studied. SLUB wins!
Great reference for a refresher or for a new comer. Really like the battery drain benchmarks and how you discuss some myths. Governors is great...love Smartass V2 with Matrix V20!!
Thanks @bedalus!! Totally deserves a sticky.
Sent from my Nexus S using xda premium
Thanks. It's in the sticky roundup!
bedalus said:
Thanks. It's in the sticky roundup!
Click to expand...
Click to collapse
Can we vote for the sticky or it is just for the moderators?
There was like ten stickies, I think veyka wants to keep everything in a list in one sticky.
I appreciate the sentiment though!
bedalus said:
There was like ten stickies, I think veyka wants to keep everything in a list in one sticky.
I appreciate the sentiment though!
Click to expand...
Click to collapse
thank you for all your work, love your kernels ... currently been using deadline/ondemand but since you say it is not best I will try smartassv2+noop.. (I've found smartass with other combinations give me a lot of lag)
BLaZeDRas said:
thank you for all your work, love your kernels ... currently been using deadline/ondemand but since you say it is not best I will try smartassv2+noop.. (I've found smartass with other combinations give me a lot of lag)
Click to expand...
Click to collapse
Ondemand for good balance between performance and battery saving. Smartassv2 for gaming. CFQ and deadline are the best governors, but deadline probably has the edge.
Thanks! Now this widens my perspective on choosing my own Governor and Scheduler Combo.
Hi all xda! I am using Gustavo_s kernel for a long time. And If you don't like stock kernel values then read this thread.
What you need:
1. A GT-I9100
2. Gustavo_s kernel,
3. Synapse (from play store)
4. Devil kernel manager (from play store)
5. A little courage
Gustavo_s kernel is a feature rich kernel but some says it is not stable [I was agree with them since chrome beta update (because chrome freezes phone only on this and all fuse bug fixed kernels)]. Here is my settings for best battery life and performance balaced! Leave unchanged I don't write any setting here.
A. Devil kernel manager settings:
CPU settings:
idle mode: idle+lpa
B. Synapse settings
1. Cpu settings
Cpu max: 1GHz
Voltage
1GHz: 1075mV
800MHz: 1000mV
500MHz: 950mV
200MHz: 850mV
2. Gpu settings
4x MSAA: CHECK
GPU 4th step freq: 400 MHz
Voltage
1st: 800mV
2nd: 850mV
3rd: 950mV
4th: 1150mV
3. Colors
If you are complaining about washed up colors or yellow tint display look at my other thread here
Yellow tint fix
4. Sound
i9100 Microphone is too silent. Boost it by 20dB
5. BLN
Force disable: CHECK
Led voltage: 2500mV
6. I/O
internal and external storage settings;
sio
1024kb
7. TWEAKS
All 650mA
8. MISC
Uncheck logcat if you don't use it.
If this settings are helpful for you please hit thanks button
Interesting thread. Good tweaks though. Please make sure to advise people that the voltages you applied will not work with all people's phones, but in most cases it should be fine.
gsstudios said:
Interesting thread. Good tweaks though. Please make sure to advise people that the voltages you applied will not work with all people's phones, but in most cases it should be fine.
Click to expand...
Click to collapse
I usally use -25mV more undervolted but this values are safe and smooth
Can I ask what ROM you are currently using?
RuffBuster said:
Can I ask what ROM you are currently using?
Click to expand...
Click to collapse
Cm11 linaro sabermod
Thnx for awesome setting..:thumbup:
---------- Post added at 09:51 PM ---------- Previous post was at 09:49 PM ----------
But gallery is not open when I check 4x MSAA..??
Did you ever got the phone freezed with these settings (apart from Chrome's bad behaviour)?
vickymbbs said:
Thnx for awesome setting..:thumbup:
---------- Post added at 09:51 PM ---------- Previous post was at 09:49 PM ----------
But gallery is not open when I check 4x MSAA..??
Click to expand...
Click to collapse
When msaa x4 is enabled, the GPU would have to work harder to render the visuals whether it is an app or a game. Some apps are not compatible with this option is ticked and that's why it is disabled by default.
thks, I will try
nhmanas said:
Gustavo_s kernel is a feature rich kernel but some says it is not stable [I was agree with them since chrome beta update (because chrome freezes phone only on this and all fuse bug fixed kernels)].
Click to expand...
Click to collapse
i don't understand. you say chrome beta freezes the complete phone on fpbug fixed kernels? and not on others? how does this happen? can you elaborate?
Lanchon said:
i don't understand. you say chrome beta freezes the complete phone on fpbug fixed kernels? and not on others? how does this happen? can you elaborate?
Click to expand...
Click to collapse
This was a long time ago mate. I wrote almost same thing on your fpbug fix thread too if you remember
-zalo said:
THANXS
Click to expand...
Click to collapse
Please don't quote unnecessary long posts.. You can just mention me.
Hey everyone! To further help old/new users of these forums understand why and how their phones can perform better or worse than some, post your AVS groups which you can find in the Info section (avs summary sub section) of the Synapse app. Here s mine...
aldoblaga said:
Hey everyone! To further help old/new users of these forums understand why and how their phones can perform better or worse than some, post your AVS groups which you can find in the Info section (avs summary sub section) of the Synapse app. Here s mine...
Click to expand...
Click to collapse
here's mine.... not sure what it all means
edwardob said:
here's mine.... not sure what it all means
Click to expand...
Click to collapse
Thanks so much for posting! Well it means some people (with higher AVS groups) can possibly overclock/undervolt better than others. For example i can apply a -75mv undervolt to all the frequencies and achieve perfect stability. Also i can overclock to the full 2.4ghz and it will reflect in everything from geekbench to antutu scores and all this while maintaining the same stock coolness of the phone and battery performance.
I m curious to see if there s anybody here with an AVS group of 10 or above and see how well they can overclock/undervolt and generally speaking perform as far as scores and battery life goes.
For example this is how far i could push with perfect stability this phone while undervolted @ -75mv - http://forum.xda-developers.com/showthread.php?p=63589738
Also the battery performance i got with 2.4ghz a57 cores and 852mhz gpu clock - http://forum.xda-developers.com/showthread.php?p=63750002
And with 1.3ghz a57 speed/420mhz gpu speed - http://forum.xda-developers.com/showthread.php?p=63851201
Mine is 12 12 12 12 13 12
alexander1995 said:
Mine is 12 12 12 12 13 12
Click to expand...
Click to collapse
OMG OMG! I WANT YOUR PHONE! ) will you exchange it for mine?
Anyway, joke aside..can you try and see how far you can undervolt/overclock and maybe some benchmarks (antutu and geekbench should be fine) and then a battery life test ?
Thanks in advance man and congrats for an awesome device!
However, it doesn't make any difference except the lower temperature. The power drain of this N920I is similar to another Note5 Which is N9200 with 7 7 7 7 8 7 avs profile. And the score of N920I which can reach 67xxx-68xxx is just a little bit higher than the one of N9200 Which is within 65xxx-67xxx
---------- Post added at 07:28 AM ---------- Previous post was at 07:25 AM ----------
I don't think it can undervolt too much. I use arter97 kernel and just undervolt 37.5mv. Haven't tried other volt
aldoblaga said:
Hey everyone! To further help old/new users of these forums understand why and how their phones can perform better or worse than some, post your AVS groups which you can find in the Info section (avs summary sub section) of the Synapse app. Here s mine...
Click to expand...
Click to collapse
Where do you find "Info" in Synapse? The only choices I have are: A57 Cluster, A53 Cluster, GPU, and Busses. I'm on Arter kernel...
First root ur device, then use an app call Root Explorer to check out /sys/kernel/debug/avs summary.
---------- Post added at 06:26 PM ---------- Previous post was at 06:25 PM ----------
ElQuixote said:
Where do you find "Info" in Synapse? The only choices I have are: A57 Cluster, A53 Cluster, GPU, and Busses. I'm on Arter kernel...
Click to expand...
Click to collapse
First root ur device, then use an app call Root Explorer to check out /sys/kernel/debug/avs summary.
alexander1995 said:
However, it doesn't make any difference except the lower temperature. The power drain of this N920I is similar to another Note5 Which is N9200 with 7 7 7 7 8 7 avs profile. And the score of N920I which can reach 67xxx-68xxx is just a little bit higher than the one of N9200 Which is within 65xxx-67xxx
---------- Post added at 07:28 AM ---------- Previous post was at 07:25 AM ----------
I don't think it can undervolt too much. I use arter97 kernel and just undervolt 37.5mv. Haven't tried other volt
Click to expand...
Click to collapse
I dunno man...i got.these.scores with a57 cores pumped at 2.4ghz both min and max... The battery score.is innacurate as i was already using the phone for an hour before starting the test..so it should have read like 10hrs and 41mins...
Avs 8 here?
Sent from my SM-N920G using Tapatalk
My asv_summary reads
big:10, LITTLE:10, INT:10, MIF:8, G3D:10, ISP:10
I've been having pretty decent success with overclocking (1.5 and 2.4 GHz) and significant undervolting (-50 mV). I could undervolt less, but I was getting some issues (freezes, rebooting) so I'm happy with where it's at now.
aldoblaga said:
I dunno man...i got.these.scores with a57 cores pumped at 2.4ghz both min and max... The battery score.is innacurate as i was already using the phone for an hour before starting the test..so it should have read like 10hrs and 41mins...
Click to expand...
Click to collapse
stock clock & fixed max freq
alexander1995 said:
stock clock & fixed max freq
Click to expand...
Click to collapse
Fixed max freq @ 2.1 or what value exactly? Good score anyway
alexander1995 said:
stock clock & fixed max freq
Click to expand...
Click to collapse
What benchmark is taht?
aldoblaga said:
Fixed max freq @ 2.1 or what value exactly? Good score anyway
Click to expand...
Click to collapse
Yeah. I didn't oc but uv using arter97 kernel.
---------- Post added at 04:51 PM ---------- Previous post was at 04:50 PM ----------
Willyman said:
What benchmark is taht?
Click to expand...
Click to collapse
PCMark. Developed by Futuremark.
I've only made small changes like I described, no fixed max frequencies for example. I get a Work score of 5700, so slightly higher than stock. I will say the Gundal Kernel makes the phone run notably smoother.
By the way, doesn't fixing the clock speed kill the battery life or make it run very hot? Surely you only do that for the benchmark right?
Willyman said:
By the way, doesn't fixing the clock speed kill the battery life or make it run very hot? Surely you only do that for the benchmark right?
Click to expand...
Click to collapse
Of course
aldoblaga said:
Fixed max freq @ 2.1 or what value exactly? Good score anyway
Click to expand...
Click to collapse
Clean up the background apps and I got this
alexander1995 said:
Clean up the background apps and I got this
Click to expand...
Click to collapse
Awesome man! Try these other tests out and let s see how.they compare...
http://forum.xda-developers.com/showthread.php?p=63903775