Well guys,
I think I found out how to only use the companion core LOL.
I use the cpu utility and have since it has been released and I know that it works well. I am able to see frequency and load per core.
I was messing around in setcpu (using motley 2.3 kernel) and when I set my maximum frequency to 475mhz only one core gets loaded. Core one in fact is the only one to show that it is working. I have ran a few benchmarks like antutu that normally show all four cores being loaded and still only one shows up.
Give it a shot. Maybe better battery life? Just thought it was interesting that we may have some control after all.
Sent from my Transformer Prime TF201 using Tapatalk
insane battery life probably but real world use might make it impractical
benefit14snake said:
Well guys,
I think I found out how to only use the companion core LOL.
I use the cpu utility and have since it has been released and I know that it works well. I am able to see frequency and load per core.
I was messing around in setcpu (using motley 2.3 kernel) and when I set my maximum frequency to 475mhz only one core gets loaded. Core one in fact is the only one to show that it is working. I have ran a few benchmarks like antutu that normally show all four cores being loaded and still only one shows up.
Give it a shot. Maybe better battery life? Just thought it was interesting that we may have some control after all.
Sent from my Transformer Prime TF201 using Tapatalk
Click to expand...
Click to collapse
Companion core for the win! Now I have the number 1 spot on antutu and the last spot haha.
{
"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"
}
Sent from my Transformer Prime TF201 using Tapatalk
benefit14snake said:
Companion core for the win! Now I have the number 1 spot on antutu and the last spot haha.
Sent from my Transformer Prime TF201 using Tapatalk
Click to expand...
Click to collapse
you sir are a boss!
I would say it is fairly impracticable. About as bad as running ics on the og Droid right? Running 620mhz @ 1 core.
Sent from my Droid Razr using Tapatalk
Interesting idea. I used antutu cpu master free and it worked fine.
After few small tests I think it's pretty usable - for twitter/facebook/etc. It's even possible to play games (defender ii) at 1core @ 475mhz, but web browsing is really really slow .
edit: Oh I got even less score in antutu - only 2048
This makes sense as the nVidia whitepaper on the Tegra 3 indicates that that the switchover to the low power core happens at 500 Mhz:
http://www.nvidia.com/content/PDF/tegra_white_papers/tegra-whitepaper-0911b.pdf
Try playing video with just the low power core -- that is supposedly one of the key use cases for the low power core since in theory, all of the heavy video decoding is supposed to be offloaded to the GPU and/or enhanced instruction sets.
NeoteriX said:
This makes sense as the nVidia whitepaper on the Tegra 3 indicates that that the switchover to the low power core happens at 500 Mhz:
http://www.nvidia.com/content/PDF/tegra_white_papers/tegra-whitepaper-0911b.pdf
Try playing video with just the low power core -- that is supposedly one of the key use cases for the low power core since in theory, all of the heavy video decoding is supposed to be offloaded to the GPU and/or enhanced instruction sets.
Click to expand...
Click to collapse
Video works great.
Just played The Big Bang Theory.
720p .mkv files through DICE player, works great!
Sent from my Transformer Prime TF201 using Tapatalk
Related
I know its real world usage that is what we go by for performance but after the new ota all my benchmarks so far are good except linpack. Linpack scores are now almost half what they were before the update. Can anyone else confirm this?
Benchmarks arent everything but as a tool they can give us hints to any problems especially when scores are reduces 50%.
Sent from my EVO 3D w/ Tapatalk
Nobody else can run linpack to confirm?
Sent from my EVO 3D w/ Tapatalk
Looks Like it who knows why I don't really care tho, phone seems fine to me, I saw top 10 and desire hd beats the EVO 3d LOL
{
"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"
}
Sent from my PG86100 using Tapatalk
I have the same problem. My linpack went from high 60s to mid 70s down to as low as 29. Usually around 40 now. Also noticed almost every test is inconsistent result.
Antutu score went from 4400 to around 3600 mainly due to float point score being greatly reduced.
I did a factory reset before the update and moved all apps and such to my SD. Once it booted up from ota it also won't pull my apps. Pics and music pulls but apps are nowhere to be found....not a fan of the update.
Sent from my PG86100 using XDA App
I know Quadrant is not a good benchmark but mine went down a little after the 2.0.x OTA as well. Was 2100. Now 2000.
Mike
My quadrant actually went down quite a bit as well. I was getting around 2200 consistantly with every one around 2450 every so often. Now I am around 2000 or so and around 2100 on the high end.
I even had one that came in at 1442 and I was like WTF!
I noticed that smartbench is pretty much the same as always.
I am a bit curious to see if anyone else is having these issues as well though. It would seem that at least one other person has noticed it. Anyone else?
It could be that they are allocating CPU resources differently now and the benchmark programs are not able to dominate one (or both) cores as much as they used to. That might not be a bad thing. I say that because, overall, the phone seems quite a bit snappier after the update. Scrolling seems smoother. Apps seem to open/close faster. Sense 3.0 just seems a little more fluid.
Mike
i got the same results with linpack. went from 60's to anywhere from 20's to 50's with the occasional 60's if i keep trying. very inconsistent. it doesn't feel slower though. on the plus side it seems my battery is lasting longer.
Whatever they did it doesn't seem to affect real world performance on my phone, just crazy that the benchmarks have dropped so dramatically.
Sent from my PG86100 using XDA App
Benchmarks dont mean a damn thing!___
R.I.P. HD2 hello EVO 3D!!
Im sorry benchmarks may not show how fast real world performance is but they can be helpfull, they can show us issues not shown. Linpack tests cpu calculations right? So after an update the score is cut in half so this could be a hint to an issue unseen by real world usage. I dont care whos faster or whatever but if i saw this happen on my gaming pc id be worried. Just sayin its worth looking into.
Sent from my EVO 3D w/ Tapatalk
Originally my phone averaged about 3250 but now I can't even get 3100. Anyone know a fix for this?
Why do you worry about the quadrant scores do much? All it does is give you an approximate score everytime.
Transmitted from a Galaxy far far away via XDA lightspeed telepathy.
geko95gek said:
Why do you worry about the quadrant scores do much? All it does is give you an approximate score everytime.
Transmitted from a Galaxy far far away via XDA lightspeed telepathy.
Click to expand...
Click to collapse
all I'm worried about is the drop in performance. It could have affected something other than my quadrant score
abraxo said:
all I'm worried about is the drop in performance. It could have affected something other than my quadrant score
Click to expand...
Click to collapse
It wouldn't matter, since quadrant is pretty meaningless nowadays. It could just have been an extra app that is running in the background. Simple as that lol
Transmitted from a Galaxy far far away via XDA lightspeed telepathy.
If you are worried about lower synthetic benchmark scores, load cm7 nightlies. I can get 5000 without any tricks out of the box with the cpu OC'd at 1.6GHz. If you don't perceive any slowness or excess battery drain and the scores themselves don't mean anything to you then don't sweat it.
abraxo said:
all I'm worried about is the drop in performance. It could have affected something other than my quadrant score
Click to expand...
Click to collapse
does it FEEL slower? does it lag? if not, then dont worry about it.
Is it still low if you run it a few times? Third run is usually the highest for me. The fact that the phone still has scores over 3,000 is incredible. Worry when it dips into the 1,000s.
In the end, Quadrant scores don't matter too much. It's all about how it feels during every day use. Does your phone lag now? I doubt it. Scores are affected by many things, including how many apps you have installed or are currently running in the background. In addition, scores can be easily manipulated!
See picture below:
{
"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"
}
nyydynasty said:
does it FEEL slower? does it lag? if not, then dont worry about it.
Click to expand...
Click to collapse
+1 on this. Unless I notice a change in something whether it's UI response or certain Apps lagging or taking longer than usual to open, I won't do anything based simply on my Quad scores. They seem way too inconsistent as it is running them back to back multiple times.
I did close all the apparently but the phone is still lagging.
So the phone is lagging? Examples of some symptoms?
If you haven't rooted or done anything to it, take it back to AT&T and let them take a look at it.
Hi everyone,
So I'm super excited to have CM9 now that the camera is working. So far it's been a great experience. I just want to make sure that I'm getting the best performance and battery life.
Here are my SetCPU settings. I was wondering if someone could look them over and let me know if there is any room for improvement.
SetCPU
Main Profile: 216 min 1300 max
Screen off profile: 216 min 750 max
Battery<50% Profile: 216 min 1000 max
I was thinking about maybe adding some profiles for CPU and Battery temperatures since my phone gets hot when I'm using GPS and I'm afraid of damaging something. Any thoughts?
Also in the undervolt tab of SetCPU I have each category set to -50mV. I'm not that well versed with undervolting so any additional tips may help.
Thanks in advance.
-D
Welcome to the Awesome Club! As for the screen off i personally have mine at 456 max and have little to no issues, the only time I have a problem is if im streaming music in background when screens off it will occasionally freeze on me. But you can deff bring that down from 1000 max, i personally run my max processor main speed at 1000 just because there isnt really a need, for me atleast, to run it at 1300 lol just kills battery quicker, heats up phone, and its beyond snappy for me at 1000. As for undervolting, i personally gave up on it, i never saw a massive difference. The highest you can UV the Tegra2 is -100mV at any given clock speed. As for temps bringing your clock speeds down will lower your temps for sure, as long as your under 50 Celsius your pretty good from what ive seen around. As for the adding more profiles, its just adding more crap running in the background lol the only one I personally run is the screen off. If i feel the phone getting too hot I manually check temps via SetCPU and if its getting to be an issue back down the processor for a few.
Instead of typing all my profiles out, I just took a screenshot instead (go go just waking up lazy ness)
{
"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've also got everything UV'd 100mv, but like w0lf215 said, I'm not 100% sure it makes a huge difference.
Sent from my CM9 Electrify.
I know undervolting can help with battery drainage (say if you undervolt by 75) but is that actually "good" per se when also having overclocked? I currently have no undervolt or overclock but I can't help wondering since many have both UV'd and OC'd.
I doubt it, when I was uv and and not even oc, my phone got real hot quickly when playing games.
Sent from my GT-I9300 using xda app-developers app
Theshawty said:
I know undervolting can help with battery drainage (say if you undervolt by 75) but is that actually "good" per se when also having overclocked? I currently have no undervolt or overclock but I can't help wondering since many have both UV'd and OC'd.
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"
}
Nice try.
The main advantage to undervolting is less heat, as we all know the more current passes through something, the hotter it gets so obviously less current = (slightly) less heat.
There are battery savings to be had but I don't know how much.
I don't think we need to talk about overclocking.
Nor sure what drugs the guy above me with the picture is on.
Sent from my GT-I9300 using xda premium
nodstuff said:
The main advantage to undervolting is less heat, as we all know the more current passes through something, the hotter it gets so obviously less current = (slightly) less heat.
There are battery savings to be had but I don't know how much.
I don't think we need to talk about overclocking.
Nor sure what drugs the guy above me with the picture is on.
Sent from my GT-I9300 using xda premium
Click to expand...
Click to collapse
I know that, but is undervolting good together with ocerclocking? I mean, if less current passes through the CPU, can it really bring out 1600 MHz? (if I were to overclock to that)
usually overclocking requires higher voltage to be stable.
when you try to push the cpu to clock faster than it is designed for, it requires more power to do so. (the voltage signal is degraded, like shouting over a waterfall, you need to shout louder to differentiate the V-high and V-low).
that is why you can undervolt the lower freqencies more than the higher one.
Not always true. I use -100mV on all steps. Perfectly fine because the voltage isnt the same across the board, but the UV is.
Sent from my GT-I9300 using xda premium
Why is it that 3Dmark score of poco is low conpared to other flagships this year ?
Distrackto said:
Why is it that 3Dmark score of poco is low conpared to other flagships this year ?
Click to expand...
Click to collapse
Which rom?
scores are low in 9. in 10 beta, antutu shows improvements in graphics and storage, but 3DMark doesn't run at all, so can't really say
Our device has hardware acceleration turend off by default while other OEMs use this in build.prop "debug.egl.hw=1" this boosts GPU performance and the results are quite noticeable.
Also our device doesn't use GPU rendering in UI which might reflect in benchmark apps.
$hivaM Manav said:
Our device has hardware acceleration turend off by default while other OEMs use this in build.prop "debug.egl.hw=1" this boosts GPU performance and the results are quite noticeable.
Also our device doesn't use GPU rendering in UI which might reflect in benchmark apps.
Click to expand...
Click to collapse
How to enable H/W acceleration? Also how to use GPU rendering?
shivy25 said:
How to enable H/W acceleration? Also how to use GPU rendering?
Click to expand...
Click to collapse
Simply edit this value in build prop "debug.egl.hw=0" to "debug.egl.hw=1" in buildprop
And for gpu rendering I would suggest leave it to software rendering because it's faster and more reliable to use in most of the apps. Search on XDA.
Also if you want to see the best results your device can do add these lines at the end of build.prop
"debug.egl.swapinterval=0" This will remove the android 60 FPS cap and improve gaming performance by 30%
"debug.gr.swapinterval=0" this will remove Device V-sync and give more framerate. (Note you'll face glitches and artaffects by using this because our display has 60hz refresh rate)
azaan477 said:
Which rom?
Click to expand...
Click to collapse
All miui roms
$hivaM Manav said:
Simply edit this value in build prop "debug.egl.hw=0" to "debug.egl.hw=1" in buildprop
And for gpu rendering I would suggest leave it to software rendering because it's faster and more reliable to use in most of the apps. Search on XDA.
Also if you want to see the best results your device can do add these lines at the end of build.prop
"debug.egl.swapinterval=0" This will remove the android 60 FPS cap and improve gaming performance by 30%
"debug.gr.swapinterval=0" this will remove Device V-sync and give more framerate. (Note you'll face glitches and artaffects by using this because our display has 60hz refresh rate)
Click to expand...
Click to collapse
Uncapping framerate shouldn't increase performance. But it WILL DEFINITELY increase heat generation and reduce battery life.
CosmicDan said:
Uncapping framerate shouldn't increase performance. But it WILL DEFINITELY increase heat generation and reduce battery life.
Click to expand...
Click to collapse
Well he wanted to see improvement in benchmark scores and yeah uncapping won't do any good because we are limited by hardware 60hrz refresh rate so no point in uncapping.
Looking at the scores, None of the MIUI based 845 phones scored at SD 845 levels but only comparable to SD 835 on 3Dmark. Except Black shark but that's completely different device.
Miui 10 beta
{
"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"
}
Hey guys, which ROM is giving the highest scores on 3DMark right now?