SD card removal results in faster frame rate - Epic 4G General

I wonder since going to ext4 or around that time, why my framerate on quadrant as really not uniform, and rather constantly up and down like a yo-yo.
That being the case, I have have really low sub 600 scores on 3d benchmark.
{
"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"
}
Unmounting the card, I seen 20 fps on the 2d, 50 fps lock on the first 3d test going through the hall and up the stairs, 70 max with the flashing light. 58 max on the planets, and 40 fps lock on dna, all resulting in 900+ scores.
I took pics of the scores with SD card mounted and unmounted.
I will try both formatting the sd first and test. And if that doesn't improve frame rate while SD card is mounted, I will move all programs to internal storage.
Edit= I'm om bonsai 1.1.3 I think it is. At 1ghz
Sent from my SPH-D700 using Tapatalk

My findings for 3d mark scores (SD mounted) -
900 on RFS (1st run).
2nd pass : 3d benchmark ONLY = 938
SD Unmounted -
1st pass : 3d benchmark ONLY = 908
side note: with SD unmounted, I noticed a lot of FPS fluctuation during benching.
Mammon's Bonsia4All v1.1.3
for comparison sake...

my score jumped about a 100 exactly from unmounting my sd card

quik99 said:
my score jumped about a 100 exactly from unmounting my sd card
Click to expand...
Click to collapse
I only got about a 30 point boost. Do you guys make sure you are at 1ghz before each test?
Sent from my ext4 Epic 4G running Bonsai4all 1.1.3!

Thanks for the testing input guys!
I'm about to test nena mark
Sent from my SPH-D700 using Tapatalk

Related

Bench Tests

Ok guys I did some benching using the Quadrant and ShootMe programs from the market. Here is my result using darkstone's froyo. I'll post more as I test them. Love to see some results from all the builds/users just for comparison. So get snappin folks! Also would be useful if you post your setup. Here's mine-->
Tmobile HD2
NRG Reference Rom 7/21
2.12 Radio
class 6 card
Darkstone's Froyo
{
"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"
}
Here's mine:
TMOUS Leo
2.10 radio
Miri's rom
stock 16GB sd
mattc build
I'll have to give that one a spin.
Sent from my HTC Desire using XDA App
Don't have a picture, but:
Quadrant score: 1256
Linpack: ~28MFLOPS
setCPU Short Bench: 300ms
T-Mobile HD2
Energy Rom 21911
2.08 Radio
Transcend 8GB microSDHC Class 6
Darkstone Froyo v1
CPU @ 1ghz Performance Mode
Tips:
-Use SetCPU and raise both minimum and maximum to 998 (highest) mhz and set it to performance mode, not on demand or governor
-Run Linpack more than a few times to get a good general idea of what your MFLOPS is
-Run Quadrant at least twice (I ran it twice). The second score is almost always higher than the first one
Quadrant score: 1278
Linpack: ~27.5MFLOPS
setCPU Short Bench: 290-310ms
setCPU Long Bench: ~480ms
T-Mobile HD2
Miri's Rom v17
2.12 Radio
Stock 16GB microSDHC Class 2
Darkstone Froyo v1
CPU @ 1ghz Performance Mode
btw drained about 2-3% in the few short minutes of testing...
but htc desire it is bad on your score ? in comparaison a htc hd2

Post your quadrant, linpack, fps2d, battery, or just any type of spec here!

Do as the title says: post your specs and represent your rom here. I'm not posting mine because I don't really care, but feel free to show us all what you've got!
p.s-and make sure to NOT post specs in any other thread other than here...
Wow, absolutely surprising. Every other thread is filled with millions of people posting their specs, yet not one person has bit the bullet here yet lol. Don't be shy, this is a flame free area lol.
Linpack 48.33
Quandrant 1540
Fps 53 stdev 4.89
Cyan latest nightly
Tried to post screen shootme incompatible file type
Sent from my PC36100 using XDA App
{
"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"
}
CM6
fps2d= 52
I will add linnpack once i actually get around to running it.
CM6 RC1
Snap v3
Setcpu: 1.228
Linpack- 49.664
Linpack: ~ 8MFlops
FPS2D: ~ avg 52
Quadrant: 289
Samsung Spica, 800 Mhz, Samdroid Kitchen, 8.13 JIT
CM6 RC1
SetCPU : 1.228
Linpack : 48.33
Quadrant : 1673
Also hold the top Linpack record right now of 132.033 MFlops on a Cupcake rom.
http://www.greenecomputing.com/apps/linpack/linpack-top-10/
apristel said:
CM6
Click to expand...
Click to collapse
Has anyone ever thought the voicemail notification looks like a face?
hannsz said:
CM6 RC1
SetCPU : 1.228
Linpack : 48.33
Quadrant : 1673
Also hold the top Linpack record right now of 132.033 MFlops on a Cupcake rom.
http://www.greenecomputing.com/apps/linpack/linpack-top-10/
Click to expand...
Click to collapse
Damn I want those scores.

Random Epic4G Quadrant Results

I am using Syndicate Frozen Rom 1.1.1 and got some stunning benchmark results.
Stock ROM i was getting around 900-1000, now that I've switched to the SFR1.1.1 (EXT4) my maximum is now 2654 which is just under what my roommate gets on his dual core android tablet.
{
"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"
}
Its called EXT4
LunaticWolf said:
Its called EXT4
Click to expand...
Click to collapse
That and acs knows how to make a mean rom lol.... best scores I ever got were on acs roms
Sent from my Evo Killer using XDA Premium App
Yea ... but doesn't 1.1.1 have something that boosts the quad scores a little anyways ...?
libsqlite.so, ext4, maybe no journaling.. hes overclocked.. this is nothing special..
chris41g said:
libsqlite.so, ext4, maybe no journaling.. hes overclocked.. this is nothing special..
Click to expand...
Click to collapse
Yea ... you can go over 9000
so this is nothing .....
It may not be anything special, but i was sincerely impressed to see the score.. maybe it just means something to me.. i'm not too experienced with custom roms and such.. sorry if this is old news.
kd0feo said:
It may not be anything special, but i was sincerely impressed to see the score.. maybe it just means something to me.. i'm not too experienced with custom roms and such.. sorry if this is old news.
Click to expand...
Click to collapse
No it's cool i like looking at them also ...
aghh i remember the Hero days it was like the little train that could.
You know whats stupid i got 2439 i was running 1400/1400 on performance on cfq
did a 2nd test same settings got 2711
3rd test on noop with performance 1400/1400 and got 2435
QUADRANT IS RANDOM NOT ACCURATE
Okay good to know, any other suggestions on what benchmark may be more accurate?

sd booster

i found that this app really speed up your sd cards
so i thought to share it keep cache value to 2048
https://market.android.com/details?id=de.mehrmann.sdbooster&feature=search_result
abhiroxx said:
i found that this app really speed up your sd cards
so i thought to share it keep cache value to 2048
https://market.android.com/details?id=de.mehrmann.sdbooster&feature=search_result
Click to expand...
Click to collapse
4096 for best results
Sent from my HTC Sensation 4G using XDA App
I think the best value depends on your hardware (phone & SD card) and ROM used. For me 3096 was the fastest without wasting more Mem on cache.
LG Optimus V + Bumblebee v2.4 + Adata 32GB Class 10 SD.
The best way is to use SD Toole to benchmark stock setting then use SD Booster to tweak and rebench till you find the best performance using the lowest cache setting.
For me its 8192!!! Fastest
O1+Mikos 6.5.8
Sent from my LG-P500 using xda premium
For me, it seems that 3096 is the ideal.
Thanks this is great 3096 works for me.
If you have rooted phone you can do sd boost and cpu tweak with ROM TOOLBOX, not free but, very, very useful.
How does this app works?
Hehe I know this app from the 1st day I rooted my phone
vlt96 said:
Hehe I know this app from the 1st day I rooted my phone
Click to expand...
Click to collapse
Hehe i know you are bumping
dhruv.always said:
Hehe i know you are bumping
Click to expand...
Click to collapse
{
"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"
}
vlt96 said:
Click to expand...
Click to collapse
Lol:-D
Sent Via Optimus One

My S6 Read and Write speeds for SD and RAM

I just got my 64GB S6 and here are the benchmarks for my fresh S6 after restoring my backup and installing the T-mobile software update that has already came out. I used A1 SD Bench to test these speeds.
I am super impressed, with these speeds, I really don't mind no micro SD slot. :good:
{
"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"
}
chillsen said:
I just got my 64GB S6 and here are the benchmarks for my fresh S6 after restoring my backup and installing the T-mobile software update that has already came out. I used A1 SD Bench to test these speeds.
I am super impressed, with these speeds, I really don't mind no micro SD slot. :good:
Click to expand...
Click to collapse
Damn those are some nice numbers! Can't wait to pick up one of these bad boys next week!
I know you get faster file movement (copying etc..), booting time along with faster programs.. but, how does this exactly affect the overall performance?
Scorpion67 said:
I know you get faster file movement (copying etc..), booting time along with faster programs.. but, how does this exactly affect the overall performance?
Click to expand...
Click to collapse
The overall effect is that everything your doing is done faster, smoother, with no hesitation. So just makes the overall experience fluid. I especially notice this speed when installing an apk. Smaller programs you hit the install button and in a blink its done, literally a blink.
Scorpion67 said:
I know you get faster file movement (copying etc..), booting time along with faster programs.. but, how does this exactly affect the overall performance?
Click to expand...
Click to collapse
You pretty much answered your own question.
Every action performed on a device requires a read, write, or both (but mostly read). The faster that occurs, the faster and smoother the performance will be.
So while it'll help with the actual "movement" of files, it also speeds up access times and the overall interactions between the device's components.
It's all about removing bottlenecks from hindering the device from doing its job.

Categories

Resources