Related
Read the thread before posting bugs or issues or questions
Table of contents:
Intro
Features/what this includes
Quirks of the kernel
Why it's labeled unstable
Disclaimer
Latest kernels
Methods of flashing
Benchmarks!
Instructions for using swap
Governor exlainations
Governor strategies
Recommended apps
Locating cpu% Eaters && other negative items towards battery life
FAQ!
Changelog
Stable/safe voltage kernels
Notes
Source
Credits
Old links, go here for other aosp roms with older video drivers & older builds
Intro: I'm scaryghoul
What this includes
It's scary!
Swap
Deadline I/o scheduler
Extreme undervolt
Overclocked & underclocked values
Sleepers disabled
Toast's new video drivers
Smartass governor
New Scary governor!
Tweaked conservative governor
and much more(All of toast's kernel)
Quirks of this kernel:
Sometimes when changing speeds with something like setcpu, they take a few seconds to actually clock to that speed, turning the screen on and off fixes it and clocks it right away though.
I would advise against using the speeds 66mhz->180mhz during real usage. It's a little laggy, but works =)
This doesn't work with sense based roms like mikshift, yes it needs to be said more than once
Why it's labeled unstable: So I don't get people poking me in the eyes with spoons if/when it freezes up. It actually works quite well, but since I heavily undervolted an undervolted kernel, it is bound to not be 100% stable(It is about 90% stable =P)
Disclaimer: What everyone else puts in kernel threads, about me not being responsible, ect.
Latest kernels
The voltages of the superUV will not work for everyone, if they don't work for you, then try the builds labeled stable
Recommended speeds 245-800mhz or 245-1ghz on scary governor && no setcpu profiles - I'll keep my current speeds in my signature.
Mirrors of kernels uploaded by VICODAN
Super SUV
http://www.mediafire.com/?47perl0d43n1nyr
Super Stable kernel
http://www.mediafire.com/?9qmtzuu35cpr8yz
Super Girly kernel
http://www.mediafire.com/?1s4z4y7bh78e04j
Methods of flashing
Joeykrim's "flashimage gui" app on the market using one of the zImage files from the zips. http://forum.xda-developers.com/showthread.php?t=1045797
Teamwin's "Kernel manager" app on the market. http://forum.xda-developers.com/showthread.php?t=1030500
Flashing via recovery using one of the zips here.
Benchmarks! - All done at 1.8ghz on cm7
Quadrant standard 1.1.7 - 3900 is the highest I've seen so far I believe.
Smartbench(2runs) - 2150/1850 - Mine were around there
Linpack(3runs) - 61, 63, 63 - These won't change much at all.
Instructions for using swap
For a swapfile do something like this.
dd if=/dev/zero of=$Swapfile bs=1048576 count=$Size
Where $Swapfile is the location of the file you want, and $Size is the amount of mb for swap you want to use.
Then type
mkswap $Path
swapon $Path
Replacing $Path/$Size with your own values of course, so if I wanted 20mb of swap I'd execute the commands
dd if=/dev/zero of=/sdcard/swapfile bs=1048576 count=20
mkswap /sdcard/swapfile
swapon /sdcard/swapfile
Governor exlainations
Toasty makes one transition to the max speed and stays there(benchmarks only usually)
Batterysave! sits at the bottom and when the cpu load increases past the threshhold it scales up to the next speed and takes another load sample and keeps doing that(best on battery life/performance ratio)
Ondemand sits at the bottom and when the cpu load increases past the threshhold, it scales ot the max speed then takes another load sample and scales down accordingly
Powersave makes one transition to the bottom speed and stays there
Smartass(Quoted from another author http://www.ziggy471.com/2010/11/07/smartass-governor-info ) - "is based on the concept of the interactive governor.
I have always agreed that in theory the way interactive works – by taking over the idle loop – is very attractive. I have never managed to tweak it so it would behave decently in real life. Smartass is a complete rewrite of the code plus more. I think its a success. Performance is on par with the “old” minmax and I think smartass is a bit more responsive. Battery life is hard to quantify precisely but it does spend much more time at the lower frequencies.
Smartass will also cap the max frequency when sleeping to 352Mhz (or if your min frequency is higher than 352 – why?! – it will cap it to your min frequency). Lets take for example the 528/176 kernel, it will sleep at 352/176. No need for sleep profiles any more!"
Scary - A new governor I wrote based on conservative with some smartass features, it scales accordingly to conservatives laws. So it will start from the bottom, take a load sample, if it's above the upthreshold, ramp up only one speed at a time, and ramp down one at a time. It will automatically cap the off screen speeds to 245Mhz, and if your min freq is higher than 245mhz, it will reset the min to 120mhz while screen is off and restore it upon screen awakening, and still scale accordingly to conservatives laws. So it spends most of its time at lower frequencies. The goal of this is to get the best battery life with decent performance. It will give the same performance as conservative right now, it will get tweaked over time.
Governor strategies
Conservative - Upthreshold 85, downthreshold 60
Smartass - Sleep max 245760Hz, Ramp up at 384000Hz at a time, up threshold 90, downthreshold 60
Scary - Upthreshold 80, down threshold 45, sleep max 245760Hz, sleep min, 122880Hz
Recommended apps =)
Nofrills - App to manage I/O scheduler, clock speeds & governors.
Adw ex or go launcher ex
Colorize widgets - Great widgets to replace the lpp ones for those converting to other launchers.
AppMonster(pro or free) - Great backup utility, automatically backs up all apps.(I like it better than TB)
GPS Status - Great application for finding satalites near you and helping get a quick lock.
Any go dev team app - Great dev team, all their apps are worth looking at.
Frequently asked questions
Question: My phone is boot looping/not booting on the released suv, what do I do?
Answer: Your phone cannot handle the super undervolted values, you will need to flash the stable undervolted kernel or girly kernel
Question: My phone's graphics are running really lagishly, why is this happening? It worked fine on version 1.4 and before.
Answer: You are likely running an outdated version of your rom not supporting the new video drivers in version 1.5+, you will need to roll back to v1.4 and before
Question: My phone is running slow on the batterysave or scary governors, what do I do?
Answer: Set your minimum speed higher, those governors spend a lot of time at the lowest values
Question: My wifi/wifitether and other stuff is broken and wasn't before I flashed your kernel
Answer: READ THE BIG BOLD STATEMENT AT THE TOP OF THE THREAD THAT SAYS THIS IS NOT FOR SENSE
Question: How do I flash this?
Answer: You probably shouldn't be flashing a kernel labeled unstable if you don't know how
Question: I'm not getting the amazing battery life/benchmark scores that others are getting
Answer: Well, every device is different, so you may not be able to score as high as some others, but you should get close at least, and the battery completely depends on usage patterns
Question: My phone can't clock to 1.8ghz, or some of the other higher values but does fine on the lower ones
Answer: Every device is different and not all can handle the same speeds/voltages, you are probably better off staying away from the higher speeds, I enabled the speeds but hope that you all are mature enough to handle them
Locating cpu% eating applications.
When in ADB shell type the following
top |grep app
or if your rom supports it htop is much better(Without grep, so just "htop")
Then look for items with high amounts of cpu% while the phone is idle.
changelog:
Release1:
Swap support enabled
Unlocked 1.6-1.8ghz
Superundervolted
Release 2(1.1):
Unlocked 3 new underclock values
Took the undervolt farther.
Release 3(1.2):
Added hw3d
Release 4(1.3):
Removed interactive governor since it only causes trouble
Overall undervolt taken farther. May be unstable for some
Release5(1.4)
Disabled sleepers
Changed kernel name
Added an alternative voltage kernel
Release 6(1.5)
Cleaned up the source
Updated to toastcfh's new video drivers, greatly improved video performance and fixed a lot of issues with aosp video. - Thanks toast =D
Release 7(1.6)
Still cleaning up source
Addeed smartass governor && tweaked it, Down thresh: 70, up: 90
Added a real kernel name
Added lulz
Added two new clock speeds && modified voltage table
Tweaked conservative governor(Should be less erradict now, so scaling less==less power consumption==Yay =D)
Release 8(1.7)
Added Scarygov 0.5
Removed userspace governor
Release 9(1.7.2)
Tweaked scarygov to 0.6, removing a bug
Release 10(1.7.3)
Fixed a lag issue on scary governor by slightly optimizing the code.
Tweaked smartass.
Release 11(1.8)
Optimized scary governor's code
Safe voltage kernels && Stable kernels(come here if you can't handle the super undervolt
NEW VIDEO DRIVERS
1.8
ScaryKernel 1.8 stable http://dl.dropbox.com/u/15373824/ScaryKernel1.8safe.zip
Girly Kernel 1.8 stable http://dl.dropbox.com/u/15373824/GirlyKernel1.8safe.zip
1.7.3
ScaryKernel v1.7.3 stable http://dl.dropbox.com/u/15373824/ScaryKernel1.7.3Stable.zip
v1.7.2
ScaryKernel v1.7.2 safe uv http://dl.dropbox.com/u/15373824/Scarykernel1.7v2Safe.zip
v1.6
[UNTESTED]1.8Ghz safe voltage no sleepers http://dl.dropbox.com/u/15373824/ScaryKernel1.6SafeJesusNoSleep.zip
v1.5
[UNTESTED]1.8Ghz safe voltage no sleepers http://dl.dropbox.com/u/15373824/Scarykernel1.5StableJesus.zip
[UNTESTED]1.5Ghz safe voltage no sleepers http://dl.dropbox.com/u/15373824/Scarykernel1.5StableNoJesus.zip
OLD VIDEO DRIVERS(1.4) - For roms that don't support the new video drivers.
1.5Ghz safe voltage kernel no sleepers http://dl.dropbox.com/u/15373824/ScarykernelSafe1.5NoSleepers.zip
1.5ghz safe voltage kernel, sleepers enabled http://dl.dropbox.com/u/15373824/ScarykernelSafe1.5Sleepers.zip
1.8ghz safe voltage kernel, no sleepers http://dl.dropbox.com/u/15373824/ScarykernelSafe1.8NoSleepers.zip
1.8ghz safe voltage kernel, sleepers enabled http://dl.dropbox.com/u/15373824/ScarykernelSafe1.8Sleepers.zip
Notes:
Nothing for now.
Here is the source code for this build. I'll try to maintain it, I have a lot of kernel sources, but this will always be the latest release code, unless I state otherwise.
https://github.com/Scaryghoul
Credits: Toastcfh/bcnice20 - For 99.8% of their kernel source.
Dogejcr/Preludedrew - For helping me get my laptop setup for kernel compiling.
Testers - For flashing this even though it says unstable :-X
JoesephMother - For helping me unlock the new lower clock values && other kernel related matters =)
x99percent - I used his middle clock speeds between the 3xx->7xx values as a base(edited voltages) as well as used his smartass source.
I just got my shifty two days ago from having the heroc. It took me a full five minutes to root the sob. This kernel clocking it to 1.8ghz is sick. Good job!
On my Gingerbreaded Evo-Shifted Premium XDA App.
Nice work again scaryghoul. Quadrant scores are very reflective of real-world performance in this case.
Awesome numbers coming off this kernel.
Quadrant = 3104
SmartBench 2011 = 2095/1887
SmartBench numbers were incredible!!!!
It was unstable for me.. 1.8 frozen my phone and it wouldn't even boot up when I tried 1.5
crump84 said:
It was unstable for me.. 1.8 frozen my phone and it wouldn't even boot up when I tried 1.5
Click to expand...
Click to collapse
Just curious but what rom?
Sent from my PG06100 using XDA App
I used 1.5 SUV no sleepers and the responsiveness of the screen lags while scrolling in adw. I flashed back to the stock #29 on N#32 and the lag is gone
Sent from my PG06100 using XDA Premium App
CrashV2010 said:
I just got my shifty two days ago from having the heroc. It took me a full five minutes to root the sob. This kernel clocking it to 1.8ghz is sick. Good job!
On my Gingerbreaded Evo-Shifted Premium XDA App.
Click to expand...
Click to collapse
Glad you can handle it.
smoothtaste said:
Nice work again scaryghoul. Quadrant scores are very reflective of real-world performance in this case.
Click to expand...
Click to collapse
Thanks =O
jjb61 said:
Awesome numbers coming off this kernel.
Quadrant = 3104
SmartBench 2011 = 2095/1887
SmartBench numbers were incredible!!!!
Click to expand...
Click to collapse
Thanks for sharing here =)
crump84 said:
It was unstable for me.. 1.8 frozen my phone and it wouldn't even boot up when I tried 1.5
Click to expand...
Click to collapse
What rom and which kernel....
threeFiftyLi said:
I used 1.5 SUV no sleepers and the responsiveness of the screen lags while scrolling in adw. I flashed back to the stock #29 on N#32 and the lag is gone
Sent from my PG06100 using XDA Premium App
Click to expand...
Click to collapse
Interesting. Never had that issue, what range of speeds are you using?
JKILO said:
Just curious but what rom?
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
Scaryghoul said:
What rom and which kernel
Click to expand...
Click to collapse
The latest Cyanogen build, tried both Release 3 1.5ghz and Release3 1.8ghz. I downloaded them last night, I may have to tried the newest release.
crump84 said:
The latest Cyanogen build, tried both Release 3 1.5ghz and Release3 1.8ghz. I downloaded them last night, I may have to tried the newest release.
Click to expand...
Click to collapse
Kernels before 1.5 aren't meant for the newest Cm.
1.5+the testing ones with the New video drivers, are the ones meant for it.
Scaryghoul said:
Kernels before 1.5 aren't meant for the newest Cm.
1.5+the testing ones with the New video drivers, are the ones meant for it.
Click to expand...
Click to collapse
Well I guess that would explain why I was having issues..
crump84 said:
Well I guess that would explain why I was having issues..
Click to expand...
Click to collapse
Version 1.5+ is meant for cm7's nightlies from 31+.
When I first tried it, I felt a little lag but after I wiped it ran better than before.
Scaryghoul said:
Interesting. Never had that issue, what range of speeds are you using?
Click to expand...
Click to collapse
Stock default 245/1200.
I'll try to wipe and start fresh when I have time. I started Nightly #27 freshly wiped but haven't wiped since.
Just flashed the 1.8 kernel with the new video drivers and its a little funky on the latest evervolv. My phone can do 1.8 stable but it is probably the video drivers acting up, gonna flash the uv kernel next
Edit: flashed release 1.5 uv 1.8ghz and it was acting up as well, on release 1.4 @ 1.8 ghz and the phone is really smooth, very fast and very stable. On a side note my performance numbers blow compared to the cm7 numbers, 1400/1400 smart bench, 48 avg linpack and like 2k on quadrant.....?
Sent from my PG06100 using XDA App
threeFiftyLi said:
Stock default 245/1200.
I'll try to wipe and start fresh when I have time. I started Nightly #27 freshly wiped but haven't wiped since.
Click to expand...
Click to collapse
Okay.
DWake014 said:
Just flashed the 1.8 kernel with the new video drivers and its a little funky on the latest evervolv. My phone can do 1.8 stable but it is probably the video drivers acting up, gonna flash the uv kernel next
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
Try one of the earlier kernels then, there might be some sort of aosp framework change required to take use of the new video drivers, I can't confirm it without hearing from toastcfh but that's why I say v1.5+ is for cm7 nightly31+.
I may be completely wrong about it though.
DWake014 said:
Just flashed the 1.8 kernel with the new video drivers and its a little funky on the latest evervolv. My phone can do 1.8 stable but it is probably the video drivers acting up, gonna flash the uv kernel next
Edit: flashed release 1.5 uv 1.8ghz and it was acting up as well, on release 1.4 @ 1.8 ghz and the phone is really smooth, very fast and very stable. On a side note my performance numbers blow compared to the cm7 numbers, 1400/1400 smart bench, 48 avg linpack and like 2k on quadrant.....?
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
Edit for your edit.
Yes, as I just said I think there might be some sort of framework change required/recompile of evervolv to support the new drivers.
Are you sure the speeds are taking? As described in the quirks, it takes a few seconds to take a speed for some, and sometimes requires screen off/on. I usually don't benchmark until linpack hits 60+
crump84 said:
The latest Cyanogen build, tried both Release 3 1.5ghz and Release3 1.8ghz. I downloaded them last night, I may have to tried the newest release.
Click to expand...
Click to collapse
I thought that was the problem
Sent from my PG06100 using XDA App
Ran some more tests, this is on Evervolv with release 1.4. Set it on 1.8/1.8 performance and:
Linpack (12 runs): avg 61, few 62s
quadrant free ( 8 runs) : avg ~2600, high was 2900
smartbench (7 runs) : avg 1800/1800, high of 1800/1900
Scaryghoul said:
Okay.
Try one of the earlier kernels then, there might be some sort of aosp framework change required to take use of the new video drivers, I can't confirm it without hearing from toastcfh but that's why I say v1.5+ is for cm7 nightly31+.
I may be completely wrong about it though.
Edit for your edit.
Yes, as I just said I think there might be some sort of framework change required/recompile of evervolv to support the new drivers.
Are you sure the speeds are taking? As described in the quirks, it takes a few seconds to take a speed for some, and sometimes requires screen off/on. I usually don't benchmark until linpack hits 60+
Click to expand...
Click to collapse
I think you're right on Evervolv cause mine from yesterday works great but my testkern.a2 is crap on there.
Nice job on the undervolting. I couldnt even boot the newest one! As always, appreciate your help with all my n00b questions.
Toasted by Nightly.
DWake014 said:
Ran some more tests, this is on Evervolv with release 1.4. Set it on 1.8/1.8 performance and:
Linpack (12 runs): avg 61, few 62s
quadrant free ( 8 runs) : avg ~2600, high was 2900
smartbench (7 runs) : avg 1800/1800, high of 1800/1900
Click to expand...
Click to collapse
Thanks for sharing =D
smoothtaste said:
I think you're right on Evervolv cause mine from yesterday works great but my testkern.a2 is crap on there.
Nice job on the undervolting. I couldnt even boot the newest one! As always, appreciate your help with all my n00b questions.
Toasted by Nightly.
Click to expand...
Click to collapse
Thanks =)
---------------------------------------------------
1.6 is out for the unstable voltage.
I'll compile the stable ones later, already made the commits to the github for what I can remember I changed
Check the changelog people =D
I have used Weapon G2x R1 and RG2X MIUI, both with the latest Faux overclocking kernel. I have used both SetCPU and CPU Master. Both apps gain super user access and seem to be functioning properly... however whenever I run the benchmark apps, the results don't seem to be any different when I have the CPU set at 1ghz or 1.5ghz.... it's the same results across both OC apps and ROMs.
The phone doesn't seem to get warm as most people report when overclocking, nor is there any stability issues.
Underclocking does seem to work.
I know quadrant scores don't mean much, but I would certainly expect there to be a difference between the scores at 1ghz and 1.5ghz. The score ranges between 2200 and 3000, but never over 3000.
There isn't any other apps installed on the phone other than what comes with the ROMs.
Am I missing something or is it an inability for my specific CPU to be overclocked?
Thanks in advance.
Try Antutu Benchmark instead. It'll show the frequency at the bottom while testing.
Sent from my amazing LG G2x running Bionix Reloaded v1.4.1
I will give it a try, thanks. I will report results after.
Thanks, the_king13, it would seem the benchmark app I was using wasn't incredibly detailed in the results.
The Antutu Benchmark app shows some pretty big performance differences when I change between the frequencies.
Thanks!
Glad to be of help. Happy overclocking!
Sent from my amazing LG G2x running Bionix Reloaded v1.4.1
As you can read from the title this thread is about benchmarks results,and decided to open this topic because is very hard to find them in this forum.
So far i,ve seen very different result with this device ,even two devices with the same firmware and no tweaks installed sometimes give a `very different result.
Also people were complaining for bad performance with the 9.4.5.22 firmware and is now happy and is getting better results with 9.4.5.26.
Well,in my case was the other way around,i`ve bought this device with 9.4.5.22 preinstalled and my benchmarks results were better than now with 9.4.5.26 (Antutu was 13.348 and now is 11562) so i wonder why so different results,this devices all have the same hardware.
Anyway,lets try to keep this topic clean and use some guidelines for posting the results,something like this:
Firmware ->> 9.4.5.26 stock or custom roms
CPU mode ->> balanced or performance
Antutu ->> your score
Specify if you are using any tweak
Offcourse you can post any benchmark you want,not only Antutu.Happy benchmarking
Inviato dal mio ASUS Transformer Pad TF700T con Tapatalk 2
These are my stock results after update.. I/O is flying. I'm downloading a torrent at the moment and I am not having any slow downs at all love my tablet
Sent from my ASUS Transformer Pad TF700T using XDA Premium HD app
.26 stock (rooted), tweaks from my thread (see the signature)
balanced & interactive governor = 12431 in AnTuTu
performance & performance governor = 13558 / 13735 in AnTuTu
It's normal that the results will vary based on many factors, but they shouldn't vary TOO MUCH (>10%?).
You can also see some of my RL Benchmark (SQLite performance) results on .26 here.
Signature: Going to butthurt ban me? Well go ahead.
Do a cold boot after .26 update
Before a cold boot AnTuTu showed IO at 4xx in Balanced, forgot to test in Performance; I am using "x" for the numbers I do not remember
After a cold boot:
Balanced:
IO is 811, total is 12501
Performance:
IO 822, total is 13517
Not tweaked, not rooted, not unlocked
while being at .21 at the Performance mode IO was 64x and the total was 1350x
while being at .22 at the Performance mode total was 1310x
So looks like .22 was a bit slower than .21 but .26 is better than both at least at IO
jdeoxys said:
Signature: Going to butthurt ban me? Well go ahead.
Click to expand...
Click to collapse
Your benchmarks are the first I have seen beat the HTC and the prime.
For some reason, I keep thinking from all the benchies I have seen over the past week, those two have beat the infinity.
I might be wrong though (usually am).
Anyways, very cool.
.26 Stock
Performance
lardo5150 said:
Your benchmarks are the first I have seen beat the HTC and the prime.
For some reason, I keep thinking from all the benchies I have seen over the past week, those two have beat the infinity.
I might be wrong though (usually am).
Anyways, very cool.
Click to expand...
Click to collapse
Btw, all these were balanced mode.
Signature: Going to butthurt ban me? Well go ahead.
jdeoxys said:
Btw, all these were balanced mode.
Signature: Going to butthurt ban me? Well go ahead.
Click to expand...
Click to collapse
You're IO is WAY higher than mine on quadrant. Are you using any build.prop tweaks?
gottahavit said:
You're IO is WAY higher than mine on quadrant. Are you using any build.prop tweaks?
Click to expand...
Click to collapse
None but the SIO (or was it CFQ, I can't remember) governor.
jdeoxys said:
None but the SIO (or was it CFQ, I can't remember) governor.
Click to expand...
Click to collapse
That's very odd, curious to see others. Sio isn't available without tweaks so i'm guessing its cfq which is now the default. But my numbers match yours except you aio is much higher.
gottahavit said:
That's very odd, curious to see others. Sio isn't available without tweaks so i'm guessing its cfq which is now the default. But my numbers match yours except you aio is much higher.
Click to expand...
Click to collapse
Can you guys post your detailed results from Antutu? I was getting around 13k, but at the moment I'm scoring under 8k. I'm thinking the difference is mostly in the CPU benchmark results since those constitute the highest numbers in the overall score, but it would be nice to be able to compare each area and narrow it down.
Wish AnTuTu would get the update out allowing proper path to SD card.
It's scores don't reflect my card but the internal storage...
Here is my score:
And
Pretoriano80 said:
So far i,ve seen very different result with this device ,even two devices with the same firmware and no tweaks installed sometimes give a `very different result.
Click to expand...
Click to collapse
This is nothing of a surprise and the reason why comparing results is pretty much useless.
You can take results as an indicator for problems with your own device (if you have very bad results).
But they are not sufficient to compare with other android devices, not even with other infinity devices.
I have tried some of the benchmarks. Non told me what driver features it used. They did not even mention if they use the multi-core.
Also they do not check what performance mode is set, probably they will never be able to do this. Also users here usually do not mention it.
You can not be sure what apps and tasks run in the background. Does the sql test disable caching?
What firmware version is installed? Do they have the same hardware?
I feel I could go on forever with this.
There are just too many factors you can not control...
jdeoxys said:
None but the SIO (or was it CFQ, I can't remember) governor.
Click to expand...
Click to collapse
These are I/O schedulers, not CPU governors, a totally different thing.
Sent from my ASUS Transformer Pad TF700T using xda app-developers app
Benchmark after cold boot
Firmware --> stock 9.4.5.26 - rooted - everything default
Governor -- > interactive
CPU mode --> balanced
RL Benchmark --> first run =58 second run = 74 third run = 68
AnTuTu --> 12514 (CPU=7670,GPU=1318,RAM=2848,I/O=678) first run
12634 (CPU=7649,GPU=1319,RAM=2920,I/O=746) second run
12497 (CPU=4766,GPU=1314,RAM=2913,I/O=804) third run
BTW,i don't worry too much about the benchmarks results and i believe that real life user experience is more accurate.
GLBenchmark was updated to version 2.5, the new HD test is pure torture for current devices. Luckily the T33 chip is still on top for most of them
http://www.anandtech.com/show/6121/glbenchmark-25-performance/1
Read it and weep..
HAHAHAHA
Do you know what hte results were before the update?
Hi everybody!
I am in a situation described somewhat by the title. My tf700 is so freakin' tweaked that the last thing I can do is grab a class 10 microSD card and do the I/O fix (not done yet). I am on CROMI 3.3 (rooted, unlocked, but this is obvious ), OC kernel by _that with the ROW addon, which came just after Clemsyns kernel, so the performance mode is even more kick-ass, as it stayed after kernel flash. Fsync disabled, lag fix applied. Yeah, pretty much everything except I/O fix. The problem is my top bemchmark result is about 6700, while I expect ~7150 (well, some users with same specs as mine get ~7150). Done right after a coldboot, with all the useless apps killed, forced GPU. I don't really care abot the result, but it seems like there is some performance potential missing, yet to be used... Would be really nice if someone would be able to point out the reason of this.
Sent from my ASUS Transformer Pad TF700T using XDA Premium HD app
post a screenshot as an attachment so some good samaritan can see the cpu / io / memory / 2d /3d figures... and which one is low...
wnerw said:
Hi everybody!
I am in a situation described somewhat by the title. My tf700 is so freakin' tweaked that the last thing I can do is grab a class 10 microSD card and do the I/O fix (not done yet). I am on CROMI 3.3 (rooted, unlocked, but this is obvious ), OC kernel by _that with the ROW addon, which came just after Clemsyns kernel, so the performance mode is even more kick-ass, as it stayed after kernel flash. Fsync disabled, lag fix applied. Yeah, pretty much everything except I/O fix. The problem is my top bemchmark result is about 6700, while I expect ~7150 (well, some users with same specs as mine get ~7150). Done right after a coldboot, with all the useless apps killed, forced GPU. I don't really care abot the result, but it seems like there is some performance potential missing, yet to be used... Would be really nice if someone would be able to point out the reason of this.
Sent from my ASUS Transformer Pad TF700T using XDA Premium HD app
Click to expand...
Click to collapse
Seems like you're not hitting 1800 OC, can you download "cpu spy" and see if you actually hit 1800 in performance mode? and make sure you set your governor to "interactive".
Yeah, haha, I can't post pictures in this forum, cause XDA Premium app says so. Neither can I post URLs as I am still below 10 posts. Anyway, I'm using interactive governor. According to cpu spy my tf700 uses every frequency with no problem.
My latest benchmark results (I updated bootloader today):
Overall - 6909
CPU - 16475
Mem - 4273
I/O - 10648
2D - 833
3D - 2318
CPU result seems quite high, ain't it? I think that 2D is the problem, though I haven't compared it yet...
Sent from my ASUS Transformer Pad TF700T using XDA Premium HD app
6400+ is great performance. Quadrant is not a very good consistent bench marker. You can run it 10 times and get scores wildly varying by 500 points. I couldn't get over 7000 for 2 days. Thought something was up and then suddenly with no changes 7200.
Don't worry about it. Does the ROM feel quick?
sbdags said:
6400+ is great performance. Quadrant is not a very good consistent bench marker. You can run it 10 times and get scores wildly varying by 500 points. I couldn't get over 7000 for 2 days. Thought something was up and then suddenly with no changes 7200.
Don't worry about it. Does the ROM feel quick?
Click to expand...
Click to collapse
Good a true words ^^^
I get low scores with Quadrant, yet I don't believe them because 3.3 CROMi is very smooth and shows no indications of stutter.
Been said a billion times "it's all all about user experience."
I recently flashed CROMI-X, and I'm loving it so far; huge improvement over stock, thanks to all involved! A quick question, though: I've been running the Quadrant benchmark, and averaging around 4400. That's a big improvement over stock, but seems quite a bit lower than any scores I see posted here; could I have messed anything up during the install, missed an important setting, ticked the wrong box, anything like that?
zarquon7 said:
I recently flashed CROMI-X, and I'm loving it so far; huge improvement over stock, thanks to all involved! A quick question, though: I've been running the Quadrant benchmark, and averaging around 4400. That's a big improvement over stock, but seems quite a bit lower than any scores I see posted here; could I have messed anything up during the install, missed an important setting, ticked the wrong box, anything like that?
Click to expand...
Click to collapse
A couple of questions here:
What Kernel did you install?
What speed do you have your CPU set at? (Assuming you installed one of the overclock kernels)
pastorbob62 said:
A couple of questions here:
What Kernel did you install?
What speed do you have your CPU set at? (Assuming you installed one of the overclock kernels)
Click to expand...
Click to collapse
_that's kernel, v4, so, no overclock...
zarquon7 said:
_that's kernel, v4, so, no overclock...
Click to expand...
Click to collapse
Yeah, if you don't have overclocking, you won't reach the benchmark results posted.
But then it is all about usability, so if you are happy, no reason to chance damaging your device by overclocking.
Sent from my TF300T using Tapatalk
zarquon7 said:
_that's kernel, v4, so, no overclock...
Click to expand...
Click to collapse
If you are content with the functionality then stick with that. But if you are experiencing lag and sluggish performance then an oc'ed kernel is the only option you have to give it a boost.
I ran CROMi-X with Hundsbauh's kernel for a year, updating with each new version. I never experienced any problems while running oc'ed and the performance was great. Hundsbauh's includes an app where you can tweak the clock speeds for both the CPU and the GPU. e.g. I set my Balanced Mode to 1500 Mhz / 484Mhz and ran that way most of the time with no issues. It was a nice balance between performance and battery life.
That said, each device is different and your results may vary, but you should be safe using CROMi-X with Hundsbauh's kernel set to a reasonable overclock. It won't let you go higher than 1700 Mhz / 520 Mhz.
pastorbob62 said:
If you are content with the functionality then stick with that. But if you are experiencing lag and sluggish performance then an oc'ed kernel is the only option you have to give it a boost.
I ran CROMi-X with Hundsbauh's kernel for a year, updating with each new version. I never experienced any problems while running oc'ed and the performance was great. Hundsbauh's includes an app where you can tweak the clock speeds for both the CPU and the GPU. e.g. I set my Balanced Mode to 1500 Mhz / 484Mhz and ran that way most of the time with no issues. It was a nice balance between performance and battery life.
That said, each device is different and your results may vary, but you should be safe using CROMi-X with Hundsbauh's kernel set to a reasonable overclock. It won't let you go higher than 1700 Mhz / 520 Mhz.
Click to expand...
Click to collapse
After my original post, I tried switching to _that v7, and saw a nice bump in performance, though I'm not sure exactly what it's overclocking it to; I may try Hundsbuah's kernel, I like the idea of tweaking the speeds myself. I had been under the impression that _that's was a bit more stable, but it seems many people use Hundsbuah's without issues.