IO R/Wspeed - LG Optimus 4X HD

I have I question for more experienced folks.
In last days I was making tests with Androbench. What I noticed that that the result of sequential/random write is faster than sequential/random read.
I think it is obviously that read speed is higher than writing speed, because while reading it is just checking state of a cell, when writing it has to change state of a cell which lasts little bit longer.
I'm using official cm11 with iodak 9.5 kernel. I used lagfix app to trim partitions - small improvement. There it is noticeable:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
I also ran it on vanir kitkat ROM, but it was much worser. But here was reading faster (actually write speed was worser). The whole ROM feels laggier than cm.
Is it just bad memory chip or is this software related stuff? I think it is impossible to be so much difference in ROMs, just because of hardware support by ROMs, so there is probably some software bug and that is why writing speed is much higher in cm.
Can somebody provide some results from other ROMs? I think I should try some other apps.
Sent from my LG-P880 using Tapatalk

I can help you with some results on cm10.1 stable - iodak v9.5
Before lagfix
1. Row - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523496&d=1390035794
After lagfix
2. Row - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523502&d=1390035826
3. Noop - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523503&d=1390035864
4. Deadline - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523504&stc=1&thumb=1&d=1390035864
5. Cfq - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523505&stc=1&thumb=1&d=1390035864
6. Sio - 2048 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523506&stc=1&thumb=1&d=1390035923
7. Row - 4096 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523507&stc=1&thumb=1&d=1390035923
8. Row - 128 buffer
http://forum.xda-developers.com/attachment.php?attachmentid=2523508&stc=1&thumb=1&d=1390035923

Related

Did Google slow down FRF83 for better stability?

just ran quadrant on my phone running Kang-O-Rama froyo 0.9b2 Sp1
and despite it being overclocked to 1.13 its still slower than what seems to be older 2.2 versions?
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
I also tried it with Paul's FRF83 ROM and got pretty much the same results.
I double checked my setcpu settings were good btw
anyone want to run quadrant tests too? could just be my phone ('hope not)
1381 here, no oc with intersectRaven Kernel, pauls r19
same here with this guy: http://www.pocketpc.ch/nexus-one-ro...stom-rom-nexus-one-htc_cime-6.html#post648317 (or is that you?)
lol, second run... 1600?
third run: 1488
fourth run: 1523
What are your linpack scores?
1386 Paul's R19 no oc.
no that wasn't me...
I was getting 1317 with full on performance setting on setcpu.
linpack scores are in the 35~36 range
not complaining though, just thought it was interesting considering the kernel's OC'ed
Try another kernel, that's quickly done. Linpack is 35-36 here too. You should get above 40 with oc
MCR r19 with stock kernel... results look great... and yeah, values are almost consistent...
watercool said:
Try another kernel, that's quickly done. Linpack is 35-36 here too. You should get above 40 with oc
Click to expand...
Click to collapse
thanks. trying wildmonk's oc kernel. getting 1432 now. I'll keep trying other kernels...

[Q] can somebody help me with performance of my port MIUI for stock kernel?

hello! can somebody help me with performance of my port MIUI for stock kernel & fix bugs?
Features :
-> based on cm7
->camera
-> wifi
-> gps...
->Suitable for daily use
->Great battery life
->Supported languages: English, Russiun,Chinese, other partly
-> working themes & lock screens (only MDPI)
-> fixed some bugs
----->bugs:
-> long time of the first boot
-> low performance :crying:
-> data transmission over usb sometimes provoke a reboot phone:crying:
-> radio
Smooth only at CPU frequency of 710 MHz or higher
Link for download http://www.mediafire.com/download/m3qmgj7he672cis/Miui_v.03_beta(eng).zip
P.S. Sorry for my bad English
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
The first boot of the phone will always be a little long. The time is for the rom to settle.
For the low performance try to decrease the size of your /system partition. And if you can unlock the boot loader try to use a custom kernel.
For a working radio you will need all the permission files and the library files from Sony. You may also need some modifications in those files too.
that's the problem that I can not unlock the boot loader therefore I ask to help with performance on STOCK kernel
I already use this ROM over month - everything works fine, but not enough free RAM ( only 50Mb) & perfomance... I did a full translation into English and Russian languages and fix some bugs.....
I have 3 ROMs for stock kernel with similar problems (LEWA OS, Mokee OS & Shendu OS)
here LEWA http://androids-team.com/flashs/lewa-stock-new-projest-woody-aydarking

[XPOSED app/module] VDcfs(Dynamic CPU frequency scaling) for galaxy s6/S6e/note5/s6e+

I made a xposed module called VDcfs which means Dynamic CPU frequency scaling. This app is more like TWDVFS and could change CPU scheduling dynamicly by detecting which foreground app is running. You can appoint your foreground apps to use different kinds of CPU governers. For example, VDcfs contains 3 governers based on interactive gonverer:
1. Performance governer list -- In this governer, A57 would be used first and the maximum frequency for A57 is up to 2.1G.
2. Powersave governer list -- In this governer, A53 would be used first, the maximum frequency for A53 is up to 1.3G, the maximum frequency for A57 is up to 1G.
3. Smooth governer list -- In this governer, A57 would be used first and the maximum frequency for A57 is up to 1.6G.
You can choose which app to use one of these 3 goveners.
PS: This app need 2 important things.
(1) After your activate this module in XPOSED and reboot your phone. Please give root permission to this app.
(2) Please turn on auto-run in tw smart manager.
{
"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"
}
File Link:
https://www.dropbox.com/s/17xzm830xr3ju04/VDcfs.apk?dl=0
You should put this on the Xposed Repo, so it's easier to update! Thanks :good:
Anybody tried this yet
is there any update, in chinese forum i see u gonna bring further update for this(in some case i cant login there), but its already so long been no update?

Big chip is doing nothing !!

hey guys,
recently I discovered brilliant app ( Intel® Performance Viewer ) which can monitor cpu core usage and freqs in real-time. As I was always concerned about my Z4's underperforming 810 chip, I decided to test cpu in different apps. So heres one - M&B WARBAND which is emulated and requires a lot of horse power. I also spawned around 200 troops to harden the task. As you can see I got 5 FPS while Big Chip (A57) is literally sleeping. Naturally, A53 is struggling at getting playable framerate. Same thing happens in Antutu, GeekBench, PPSSPP, and in every power demanding apps( games, emulators and etc).
WTF is wrong with 810 CPU guys? or should I blame Sony for trash governors which can't figure out when they should activate all cores?
{
"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"
}
Just hope they fix this in marshmallow.
Sent from my E6533 using XDA Free mobile app
Can anyone test this again in MM? Would be greatly appreciated.
neestea said:
Can anyone test this again in MM? Would be greatly appreciated.
Click to expand...
Click to collapse
It's functioning much much better. There's another thread that shows benchmarks.
Sent from my E6533 using XDA Free mobile app
big.Little on this device runs HMP
The most powerful use model of big.LITTLE architecture is heterogeneous multi-processing (HMP), which enables the use of all physical cores at the same time. Threads with high priority or computational intensity can in this case be allocated to the "big" cores while threads with less priority or less computational intensity, such as background tasks, can be performed by the "LITTLE" cores.[10][11] source from Wikipedia.

[CLOSED (duplicate)] [MODULE] GPU Turbo Boost

< mod edit: duplicate thread. Please continue the discussion here: https://forum.xda-developers.com/apps/magisk/module-gpu-turbo-boost-t3808541 >
{
"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"
}
Introduction:​This module enables GPU Turbo tweak to increase the phone’s GPU power by up to 75% but at the same time reduces power consumption by up to 25% seamlessly.
Also this module increases gaming experience, video acceleration, performance tuning, graphic boost and trigger gpu/cpu to boost your device to the highest level.
Requirements:​- Unlocked Devices with Custom Recovery
- Magisk Installed v15.0 and up
- Knowledge about Magisk Modules
- A little piece of brain (if necessary)
Changelogs:​- Initial Release v1.0
Downloads:​- View attachment 4534963
Contributors:​- @EmperorEye
Credits:​- @topjohnwu
- Nomelas (for his reset prop script)
Github Repo:​- https://github.com/EmperorEye1993/GPU-Turbo-Boost
Question & Answer:​Q: What this module does?
A: Read the introduction and module README.md file.
Q: Is this module available to all devices?
A: Of Course, just read the requirements.
Q: This module similar to Huawei GPU Turbo Technology?
A: Some codes, frequencies and values. but I used native reset prop tweaks for difference purposes.
Created 2018-06-25
Your simple thanks is highly appreciated.

Categories

Resources