Database Users

Informative Benchmarking article

There is a very nice article on the performance of high end Android phones at
http://www.zdnet.com/blog/btl/benchmark-showdown-high-end-android-smartphones/39206

Good read but old news
Sent from my SGH-T959 using XDA App

Resume
See Attached pictures. The major blow is that there will be no joy in upgrading stock to A2.2
The memory speed in SGS is superior to the competitors. The processor is basically the same ARM 1GHz. But the overall performance lags 4 to 6 times depending on how you account performance. Respectively, the battery drain will be inferior and the lags continue to drag on.
Most likely, the infamous time lag issue that is related to "inefficient" memory management plays the role. There was no official Samsung reaction upon the issue. Considering how memory management is done now (the device -most likely- minimizes memory triggering, doesn't close unused apps despite significant battery drain [200+ hours after boot up, wifi, 2G cell, no apps touched vs. less than 24h if you use apps along the way]) and the deliberate ignorance of the situation on Samsung's side, one can conclude that there is a reason. It could be a short life time for the memory cells, i.e. trade off speed vs number of write cycles. Or something else of the sorts.
http://forum.xda-developers.com/attachment.php?attachmentid=402289&stc=1&d=1284656758
http://forum.xda-developers.com/attachment.php?attachmentid=402290&stc=1&d=1284656758

I agree, it's a decent article that helps explain some of what composes each benchmark. I do, however, believe it missed one key point regarding the 2.2 beta benchmarks. From everything I've read, the Froyo beta ROM hasn't even enabled (muchless optimized) the Dalvik JIT compiler that accounts for the huge differences in CPU performance shown on Page 2. Essentially, the graph shouldn't show the 'Froyo / beta ROM' result in green quite yet. We're also not shown the great benchmarks resulting from lagfixes. I don't know if that's also true with the Froyo Legend, but it certainly discounts the full potential of our Galaxy S phones.
Seeing how well the Galaxy S ranks otherwise leaves me thinking it remains the best Android hardware available today. And with all the positive feedback I'm hearing about software fixing both the lag and GPS issue, I'm quite excited. But only time will tell and I, for one, can wait for the devs to finish tweaking and testing things for us. Now if there were a port of Quadrant for jailbroken iPhones, we could compare those too. (There are a few Linpack attempts but they don't seem to be reliable.)

Error all over the place
First of all I didn't like this article at all.
Also, someone should notify them that their conclusion is wrong. The Hummingbird is 45nm, not 65nm

Yeah, I think it's weak they're comparing a Galaxy with an unofficial Froyo install versus ones that have had matured Froyo for a while. And it's their only data point for claiming that the Snapdragon cores benefit from Froyo more than the OMAPs or Hummingbirds... kind of a weak claim before we have some more data.

Patience,and time, will show what the Vibrant will perform much better with actual 2.2 vs. the builds used in the review.

RMA'd Nexus 4 is much slower than my old Nexus 4 + much yellower screen

I had my Nexus 4 for 15 days, the GPS stopped working, and I called in for an RMA. Got it today and started comparing; the new phone's screen is extremely yellow compared to the old phone's and it's very noticeable right off the bat. The new phone is also very slow compared to the old phone according to Antutu benchmarks (12000-15000 compared to 16000-19000). The following are screenshots that indicate the screen tint and the benchmarks:
imgur.com/a/q3KWM
Any help would be much appreciated!
Edit: The phone itself in daily use is also very stutter-y at times and it's very noticeable.

Toyotaspeed said:
I had my Nexus 4 for 15 days, the GPS stopped working, and I called in for an RMA. Got it today and started comparing; the new phone's screen is extremely yellow compared to the old phone's and it's very noticeable right off the bat. The new phone is also very slow compared to the old phone according to Antutu benchmarks (12000-15000 compared to 16000-19000). The following are screenshots that indicate the screen tint and the benchmarks:
imgur.com/a/q3KWM
Any help would be much appreciated!
Click to expand...
Click to collapse
Your thread title is very misleading. I thought you were talking about the phone performing slowwwww.
Firs off, low benchmarks does not justify a device been slow or quick. That all depends on a the CPU. Second, no two exact phone will put out the same benchmark. If a mother gives birth to two twins, you don't expect them to do the same exact things together all the time 24/7/365 days a week. They look the same but there are differences that put them apart.
Benchmarks are made to show off the awesome score the device gets. Benchmarks do not affect real life tasks when you use your phone.

Tasks themselves are also slow. The UI is stutter-y at times and side-by-side, the new phone launches apps and loads videos, etc slower, sometimes with a difference of up to 1-2 seconds. Equating factory-built hardware to genetic make-up isn't exactly a very good analogy. There is a pretty big difference in performance in both benchmarks and actual use.

scream4cheese said:
Your thread title is very misleading. I thought you were talking about the phone performing slowwwww.
Firs off, low benchmarks does not justify a device been slow or quick. That all depends on a the CPU. Second, no two exact phone will put out the same benchmark. If a mother gives birth to two twins, you don't expect them to do the same exact things together all the time 24/7/365 days a week. They look the same but there are differences that put them apart.
Benchmarks are made to show off the awesome score the device gets. Benchmarks do not affect real life tasks when you use your phone.
Click to expand...
Click to collapse
Not to mention that the same device may have different results in consecutive tests...

install a calibrated kernel, this one for instance: http://goo.im/devs/paranoidandroid/roms/mako/pa_mako-Kernel422_CalibratedContrasty.zip
just wash over in cwm/twrp then reboot
antutu behaves weird i wouldnt bother with it. theres so many factors that can screw with its results, apps running in the background fetching data or whatever. jerky UI shouldn't be but then again, might be the same app thats trashing your stats. try installing a rom as well for all its worth.

that yellowing is really bad and uneven. supposedly, it's the glue and sometimes goes away with some use. If it doesn't, definitely rma that.

LG must use thermal compound

Guys, look at this
This is pathetic results for SD800. And the reason of so low bench results is absence of thermal pad on the SoC.
This was the case with first Nexus 4 http://forum.xda-developers.com/nexus-4/general/hardmod-nexus-4-investigating-thermal-t2144652, the same is with Nexus 5 http://forum.xda-developers.com/goo...de-nexus-5s-throttling-solved-adding-t2627714 and the same goes to LGG2.
I mean, when I spend $600+ for expensive phone I want it perform better than competitors' ones, I mean Xperia Z1, Note 3, Xperia Z Ultra. I what my bang for the buck
What do you think? Would you buy G Pro 2 or LG G3 if they will lack thermal grease too?

Benchmarks mean nothing, the phone runs fine and quickly, games play fine at good frame rates, so its pointless doing this.
Transferring heat away from a components thats designed to handle it to a component that isnt is a bad idea. You may not think the screen is getting heated, but there a layers of very insulating glass between you and the LCD, you wont be aware just how warm it is. Heat affects LCDs, in time you might end up with a dark spot where the CPU is.

Billy Madison said:
Guys, look at this
This is pathetic results for SD800. And the reason of so low bench results is absence of thermal pad on the SoC.
This was the case with first Nexus 4 http://forum.xda-developers.com/nexus-4/general/hardmod-nexus-4-investigating-thermal-t2144652, the same is with Nexus 5 http://forum.xda-developers.com/goo...de-nexus-5s-throttling-solved-adding-t2627714 and the same goes to LGG2.
I mean, when I spend $600+ for expensive phone I want it perform better than competitors' ones, I mean Xperia Z1, Note 3, Xperia Z Ultra. I what my bang for the buck
What do you think? Would you buy G Pro 2 or LG G3 if they will lack thermal grease too?
Click to expand...
Click to collapse
What exactly is the "bang for your buck" that you require? What does the G2 not do for you that the Note 3 or Z1 does that is so crucial to daily life?
Because if it is just benchmarks, then as mentioned above, they mean nothing.

Billy Madison said:
Guys, look at this
This is pathetic results for SD800. And the reason of so low bench results is absence of thermal pad on the SoC.
This was the case with first Nexus 4 http://forum.xda-developers.com/nexus-4/general/hardmod-nexus-4-investigating-thermal-t2144652, the same is with Nexus 5 http://forum.xda-developers.com/goo...de-nexus-5s-throttling-solved-adding-t2627714 and the same goes to LGG2.
I mean, when I spend $600+ for expensive phone I want it perform better than competitors' ones, I mean Xperia Z1, Note 3, Xperia Z Ultra. I what my bang for the buck
What do you think? Would you buy G Pro 2 or LG G3 if they will lack thermal grease too?
Click to expand...
Click to collapse
not long ago the g2 rocked the benchmarks mwahhaha so there with your thermal grease

bachera said:
not long ago the g2 rocked the benchmarks mwahhaha so there with your thermal grease
Click to expand...
Click to collapse
the kk update killed the bench score, i think mine dropped like 10k.

Benchmarks are the worse thing to ever happen to android. You would think that the moto x with its "out of date" specs still running smoother then the s5 would prove once and for all that benchmark scores are pointless. I don't care if the note 3 or s5 put up record breaking omg wtf fml scores...it still lags and mine doesn't! Lol....
Sent from my VS980 4G
---------- Post added at 06:37 AM ---------- Previous post was at 06:30 AM ----------
To add to this.... lets say you bought 2 cars. Both cost the same, one car says 700 hp, 0-60 in 3 seconds. The second car has 300 hp, 0-60 in 6 seconds. But once you get on the road, the 300 hp drives better, feels faster in every way and gets much better gas mileage. There's what's on paper, and there's real life.
Sent from my VS980 4G

bweN diorD said:
the kk update killed the bench score, i think mine dropped like 10k.
Click to expand...
Click to collapse
but user experience increased a lot right? ui flluidness/ opening of apps?

mgolder said:
Benchmark d*ck swinging - The rankings equivalent of being the moron who still says 'FIRST' on a comments section
Click to expand...
Click to collapse
lol, I dont mind good benchmarks, there just arent any on android that measure fluidness/real life snappyness.
First! after the guy who doesnt like firsts.

bachera said:
but user experience increased a lot right? ui flluidness/ opening of apps?
Click to expand...
Click to collapse
i dont see it. my phone was snappy and fluid before the update, its the same now.
as for the benchmark comments. i agree to a degree that they dont mean a lot, but a 25-35% drop seems a bit extreme for an update.
seems like they boosted the benches to boost sales when it was new.

The fluidity and the battery are awesome. I love everything on that phone. Even if LG doesn't keep it updated it will still be one great phone that i own/ed...

bodom_hc said:
Its pathetic you ACTUALLY care about benchmarks, lol.
Click to expand...
Click to collapse
I don't care bout benches, I care about what I paid for, and I use, as everyone else does, benches to prove that my item works as intended. Did you hear about lags when doing spherical panoramas? when phone overheats and process too long, longer than it must? It's the same story. Did you hear about "slow focus" in G2? Everybody did. Maybe it partly is because of early frequency drop, maybe not so.
And not only I care for benchmarks, look there http://blog.gsmarena.com/lg-g2-jelly-bean-vs-kitkat-benchmark-battle/

Billy Madison said:
I don't care bout benches, I care about what I paid for, and I use, as everyone else does, benches to prove that my item works as intended. Did you hear about lags when doing spherical panoramas? when phone overheats and process too long, longer than it must? It's the same story. Did you hear about "slow focus" in G2? Everybody did. Maybe it partly is because of early frequency drop, maybe not so.
And not only I care for benchmarks, look there http://blog.gsmarena.com/lg-g2-jelly-bean-vs-kitkat-benchmark-battle/
Click to expand...
Click to collapse
But you didn't pay for a device that guarantees you any particular score in artificially inflated benchmark tests though. No manufacturer will launch a device guaranteeing someone a set score all the time. So what is it that you think you should have that you haven't got now? I thin that is what we can't understand.

mgolder said:
But you didn't pay for a device that guarantees you any particular score in artificially inflated benchmark tests though. No manufacturer will launch a device guaranteeing someone a set score all the time. So what is it that you think you should have that you haven't got now? I thin that is what we can't understand.
Click to expand...
Click to collapse
Man, I want it to not lag earlyer than do another smartphones. Did you hear of lags when playing Spiderman 2? I've heard that G2 is not capable of playing this recent new game smoothly, as device with SD800 should. OK I don't play Spiderman2 and did not by to play benchmarks, I only use benchmarks as examples that internally G2, Nexus 4, Nexus 5 do not constructed properly. And as I did http://forum.xda-developers.com/showthread.php?t=2299888 when I complained of improper color temperatures of G Pro LG reacted and made G2 display perfectly calibrated and on par with HTC One and iPhone 5S. So I think that in future G3 they'll do something with SoC cooling and improve performance when gaming for long time

Billy Madison said:
And not only I care for benchmarks, look there http://blog.gsmarena.com/lg-g2-jelly-bean-vs-kitkat-benchmark-battle/
Click to expand...
Click to collapse
These things have nothing to do with benchmarks:
Did you hear about lags when doing spherical panoramas?Overheating can be many things, I have yet to find a phone that doesnt overheat sometimes. Usually a clean instal fixes this nicely. Its Android I guess and faulty install scenarios.
It's the same story. Did you hear about "slow focus" in G2? -> slow focus is the automatic focus, press on screen where you want focus and its almost instant. Google camera is pretty good. there are other settings that can tweak this.
That said, benchmarks in my opinion do nothing good for true user experience.
I do believe Android needs to step up on the user experience game as my comments above prove enough that this device, I say all android devices, need to much customization to be perfect. All these workarounds To fix these real issues that some users find anoying, not higher benchmarks are needed.

lol, and does kitkat score less in that? didnt see you post scores once with that nor found your arguments on it.Stop trolling yourself and post some proof then.
so my arguments from that standingpoint still hold.
I also couldnt find it featured on any major android website, wonder why.
the one benchmarktool I dont know, its only got 33 reviews, hardly antutu level. wonder if its representative.
This could be a benchmark that actually does measure some important stuff I am talking about. Interesting. thanks for sharing
omg the downloading is soo slow wow

ok ran the benchmark. but I am not impressed as it benchmarks with its own gui, not the actual phones gui. Also it doesnt benchmark the panorama shots or any camera shots for that matter.
I guess I wasnt wrong after all... sad was hoping big on this benchmarking tool to finally have a representative view of a gui from a user point of view.

People, so upsest with benchmarks, its not like its a gaming computer/server where you squeeze every bit of performance regardless of powerdraw etc, its a phone lol, and its smoother than before KitKat. Xperia Z is a good example of benchmarks vs sucking in real life ex horrible battery life....

bachera said:
ok ran the benchmark. but I am not impressed as it benchmarks with its own gui, not the actual phones gui. Also it doesnt benchmark the panorama shots or any camera shots for that matter.
I guess I wasnt wrong after all... sad was hoping big on this benchmarking tool to finally have a representative view of a gui from a user point of view.
Click to expand...
Click to collapse
Well, from the user point of view you should your Gallery scroll, Zoom scroll, list scroll results. The gallery scroll is what fubar in LG's UI. It scores around 35-37FPS, right? But on the Nexus it scores 50-55 fps or near. All other scrolls on LG are pretty near 60 fps, so as on every other smartphone and all right with them. But gallery scroll is bad. Did you use Tapatalk ever? Have you noticed when you try to swipe from right to left to open next page the tapatalk just stagger, hangs for .5sec, and then completes the move? It shouldn't, scroll must be buttery as is in HTC phones
abhinav.tella said:
People, so upsest with benchmarks, its not like its a gaming computer/server where you squeeze every bit of performance regardless of powerdraw etc.
Click to expand...
Click to collapse
Check my hwbot prime score it is 1688 and compare to Nexus 5 scores that has same MSM8974A packed in LG box. 2923 points. pretty 50% increase just because of software, huh. Or the next result , 3075

Billy Madison said:
Well, from the user point of view you should your Gallery scroll, Zoom scroll, list scroll results. The gallery scroll is what fubar in LG's UI. It scores around 35-37FPS, right? But on the Nexus it scores 50-55 fps or near. All other scrolls on LG are pretty near 60 fps, so as on every other smartphone and all right with them. But gallery scroll is bad. Did you use Tapatalk ever? Have you noticed when you try to swipe from right to left to open next page the tapatalk just stagger, hangs for .5sec, and then completes the move? It shouldn't, scroll must be buttery as is in HTC phones
Check my hwbot prime score it is 1688 and compare to Nexus 5 scores that has same MSM8974A packed in LG box. 2923 points. pretty 50% increase just because of software, huh. Or the next result , 3075
Click to expand...
Click to collapse
OK, since you mentioned it I did that and got 2745 in HBOT Prime which is not too far off from the N5.
As for HTC I agree, since the One X days they have gone pretty far, they are probably best Android manufacturer at Battery/Performance optimization IMO, and Sony is the worst.
My HTC Butterfly (Int DNA) with 2000 mAh battery lasts longer than the Xperia Z (2330 mAh) per charge. Sony Exmor cameras in iPhone and Galaxy S4 perform better than the Z1. At least LG aren't like Sony.

abhinav.tella said:
OK, since you mentioned it I did that and got 2745 in HBOT Prime which is not too far off from the N5.
Click to expand...
Click to collapse
So, well, at least it is for me to think over, is it something wrong with my unit or its just my latest ROM with knock code has bad optimization, and it will be improved in future. Or, maybe, my PVS of 3 has to do with too early heating of SoC than yours CPU. Maybe you have PVS of 0 or 1

Let's See Exactly How QHD Affects Performance

These tests were run on an LG G3 VS985. Screen resolution was changed using the Resolution Changer Pro app.
Resolutions tested were:
* 2560x1440 @ 540ppi
* 1920x1080 @ 360ppi
* 1280x720 @ 240ppi
Here are my results
**Conclusion**
QHD results in approximately a 15% hit of graphics performance on the same hardware vs a 1080p resolution.
Also notable is that in AnTuTu, even at 720p resolution, the LG G3 scores less than the LG G2. At QHD, it dips below the Galaxy S4. My guess would be that this is because of LG's power-saving features preventing the SoC from being used to its full ability.
Also of note, I literally cannot tell the difference between 1080p and QHD on this display. Even 720p still looks really good, but LG's UI has some scaling issues.

well did you test it on battery?

We already knew that a QHD display with the Snapdragon 801 was going to give you less performance than a 1080p display with a Snapdragon 800 chip. This is old news and doesn't tell us anything we didn't already know
The real question is if running the device on 1080p/720p give you more battery life than 1440p. And if so, how much more

mgbotoe said:
well did you test it on battery?
Click to expand...
Click to collapse
Plugged in to USB. Didn't want heat to be an issue by charging it at the full 1.8A

Benchmarks are a joke!
Here are my antutu scores on day 3 of owning my phone. non-root, completely stock!
Just goes to show what devs have been saying for years, benchmarks dont mean crap! lol

xallinvaynex said:
Here are my antutu scores on day 3 of owning my phone. non-root, completely stock!
Just goes to show what devs have been saying for years, benchmarks dont mean crap! lol
Click to expand...
Click to collapse
I agree, my Antutu scores are about 34,000 too. lol

thanks for the effort.

xallinvaynex said:
Here are my antutu scores on day 3 of owning my phone. non-root, completely stock!
Just goes to show what devs have been saying for years, benchmarks dont mean crap! lol
Click to expand...
Click to collapse
nah they definitely do mean something. Have you tried AnTuTu X? It prevents the clock speed from changing when it detects a benchmark running (aka "benchmark cheating", although I'm not too fond of calling it that for reasons), so it gets exactly the performance other apps get. Pretty sure LG just disables thermal throttling during most benchmarks, so 99% of the time you'll get the performance that AnTuTu X says you have rather than what regular AnTuTu says.

Turn off knock on and a few useless apps and you can make consistent benchmarks like this. This is pure stock, no root and not my highest run. Im even running a few widgets in the background.

conan1600 said:
Turn off knock on and a few useless apps and you can make consistent benchmarks like this. This is pure stock, no root and not my highest run. Im even running a few widgets in the background.
Click to expand...
Click to collapse
Who would turn off knock on though? Its pretty much the whole reason the buttons are on the back.

Osseon said:
Who would turn off knock on though? Its pretty much the whole reason the buttons are on the back.
Click to expand...
Click to collapse
I disagree. I like the back button placement. On side buttons I inadvertently always hit volume or power when using my phone. I would, and do, disable knock on as I do not find it useful, however knock on is a small thing, even with it enabled you will pull 35 to 36k antutu scores with proper settings, of course knock on will eat your battery.
I do not multi task on a small phone, there just isnt enough screen, so I set up a toggle for turning it on and off, I keep it off unless I need it which means it will never be on.
From previous lg phones I know quick remote will chew through battery and cpu if left on so I simply toggle it off until I want to use it. I do use that a lot.
I do not find google now helpful so it is off. I do find google search useful so it is on and ok google is enabled.
I don't use 3 billion widgets but I do have a weather widget and I run sync 24/7 with many accounts such as FB, DB, Etc. With just what ive stated even with knock on enabled you should pull 35k which puts this phone above everthing except the m8 on antutu x. Disable sync, uninstall cpu rape ware and your battery life and benchmarks will be even higher. Currently I have 65% battery left with 4 hours screen on time. Wifi enabled, low signal area. My way works for me.

As an avid gamer. I definitely see the difference between All 3 resolutions. You must test fairly. Download same video in all 3 different resolutions.

conan1600 said:
Disable sync, uninstall cpu rape ware and your battery life and benchmarks will be even higher. Currently I have 65% battery left with 4 hours screen on time. Wifi enabled, low signal area. My way works for me.
Click to expand...
Click to collapse
What cpu rape ware have you uninstalled? Also do you use greenify or similar apps? 4 hours screen time seems impresssive if you have 65% left.
However I must say this phone sleeps like a baby, 9 hours since I'd pulled it out of the plug. 1% discharge wtf.

scy1192 said:
Plugged in to USB. Didn't want heat to be an issue by charging it at the full 1.8A
Click to expand...
Click to collapse
So you did the test wrong and may dump your results into trash can. Tests should be performed without charging connected for heating issues to make less impact on results and phone should be cooled before each test

Billy Madison said:
So you did the test wrong and may dump your results into trash can. Tests should be performed without charging connected for heating issues to make less impact on results and phone should be cooled before each test
Click to expand...
Click to collapse
are you always such a pleasant person?

conan1600 said:
Turn off knock on and a few useless apps and you can make consistent benchmarks like this. This is pure stock, no root and not my highest run. Im even running a few widgets in the background.
Click to expand...
Click to collapse
LG cheats at AnTuTu. You need to use AnTuTu X

scy1192 said:
LG cheats at AnTuTu. You need to use AnTuTu X
Click to expand...
Click to collapse
That screen shot is antutu x. Look at the top of the ss
---------- Post added at 01:33 PM ---------- Previous post was at 01:32 PM ----------
danny_and_2 said:
What cpu rape ware have you uninstalled? Also do you use greenify or similar apps? 4 hours screen time seems impresssive if you have 65% left.
However I must say this phone sleeps like a baby, 9 hours since I'd pulled it out of the plug. 1% discharge wtf.
Click to expand...
Click to collapse
Ill do a whole battery thread after im done playing.

Enddo said:
We already knew that a QHD display with the Snapdragon 801 was going to give you less performance than a 1080p display with a Snapdragon 800 chip. This is old news and doesn't tell us anything we didn't already know
The real question is if running the device on 1080p/720p give you more battery life than 1440p. And if so, how much more
Click to expand...
Click to collapse
If mainly light use like calls, texting and local video playback it will not much, but anything like web and any games that try push to each pixel will. Gameloft is a good example, since usually have terrible scaling setup (the games behave at whatever the native res).
The advantage is for whatever uses the gpu much. A reason the UI is a little slow (not much IMO) is it is true Qhd graphics and that causes the gpu to sweat some. The mean of the 801 was designed for 1080p. The 805 is the first chip for Qhd.
---------- Post added at 06:56 PM ---------- Previous post was at 06:47 PM ----------
Anandtech and other reviews that actually do thorough tests pointed out for light use the G3 has the same or better battery life as current devices. The killer is apps like a lot of Web content and games.
SOT is irrelevant unless itemized based on actual use, brightness, what radios are on and the signal reception.

Change your nexus 6 Resolution too 1080x1920

I found this on YouTube and its a pretty good mod. You don't even see any difference.
Boost Nexus 6 Performance & Battery Life by Lowering the Resolution [How...: http://youtu.be/9HREkRpPVa4

I'll take the hit on battery life for a higher resolution. That's part of the reason why this phone is so great.

dave2metz said:
I'll take the hit on battery life for a higher resolution. That's part of the reason why this phone is so great.
Click to expand...
Click to collapse
I feel you..... but serious tho you can't see no difference at all.

Sounds interesting, might give this a try.

Have you seen a noticeable difference in battery life? If this would give an extra hour or two at least, I could see doing it. If it's only a few extra mins, maybe not. Guess it can't hurt to try.

Nothing for battery time but here are some AnTuTu tests.
2d and 3d results, Nexus 6 stock, rooted, unencrypted 5.01.
Ran AnTuTu before making the change...
2d - 1596
3d - 16910 (2560x1440)
Made the change and ran again...
2d - 1636
3d - 20004 (1920x1080)
Changed back to stock and ran again...
2d - 1636
3d - 16303 (2560x1440)

I suspect that it would actually be more clear at 1280x720. Reason is that this is precisely the half-way point, which means that one "virtual" pixel takes up 2pixelsX2pixels on the physical screen. At 1920x1080, you're somewhere in between, which means that pixel boundaries won't line up to real pixels.
1280x720 also is the "full" resolution for one direction of the RGBG subpixel matrix.

tcrews said:
Nothing for battery time but here are some AnTuTu tests.
2d and 3d results, Nexus 6 stock, rooted, unencrypted 5.01.
Ran AnTuTu before making the change...
2d - 1596
3d - 16910 (2560x1440)
Made the change and ran again...
2d - 1636
3d - 20004 (1920x1080)
Changed back to stock and ran again...
2d - 1636
3d - 16303 (2560x1440)
Click to expand...
Click to collapse
That's sad. I got 51590 on antutu, stock decrypted

wmfreak said:
That's sad. I got 51590 on antutu, stock decrypted
Click to expand...
Click to collapse
Not really.
I've gotten 52574, 52140 and 52083 all stock, decrypted for the full score just randomly running AnTuTu full test.
I just pulled out my phone today, ran the test and reported the 2d and 3d numbers only.
My full scores today were.
Initial - 51180
change - 54067
back to stock - 49606
These were all run back to back and the phone was getting warmer so that may be why the score dropped as much as it did for the final run. All runs previous to today have been in the 52,000+ range.

I changed mine to 1280x720, with a density of 320. It looks pretty good, with some minor graphical issues. At normal resolution 3d is around 17k.
2d: 1629
3d: 21205

I'll wait for your guys testing on battery life

Nice. gotta try it.

When my phone is sitting idle my GPU spends most of its time at 240 MHz, instead of 300 MHz at stock resolution. I don't know that this will save much battery as our 805's don't struggle with 1440p like the snap dragon 801 does. My wife's S5 would probably benefit from this more as it is fairly easy to bog down.
I'm having video player stuttering and pausing while trying to play 4k video that I recorded with the phone. I also noticed while navigating today that the phone didn't get as hot as it normally did, but there's a whole lot of variables that can affect that. Just some observations, I'll see how this works out tomorrow as far as battery life on a typical day.

No battery life yet eh?
Sent from my Nexus 6 using XDA Free mobile app

From the poster of the video
Anything's possible with root! But to be clear, this doesn't change the fact that your phone still has to power the extra pixels (because they're still physically there), it only helps with your GPU in that it doesn't have to process the extra pixels. But less processing generally equals less battery drain and smoother performance, and every little bit helps.
So basically sounds like battery life gains are going be very marginal at best.

Device looks like crappy with 1080p density 420. Icons looks like samsung s3. If you want that, go buy a S3

fethi2 said:
Device looks like crappy with 1080p density 420. Icons looks like samsung s3. If you want that, go buy a S3
Click to expand...
Click to collapse
I disagree with your statement, I don't see a difference at all

venturizhou said:
From the poster of the video
Anything's possible with root! But to be clear, this doesn't change the fact that your phone still has to power the extra pixels (because they're still physically there), it only helps with your GPU in that it doesn't have to process the extra pixels. But less processing generally equals less battery drain and smoother performance, and every little bit helps.
So basically sounds like battery life gains are going be very marginal at best.
Click to expand...
Click to collapse
Pure snake oil. Changing resolution does nothing. Your display/GPU is still having to draw all images at its native resolution regardless of what you set it to. Unless you picture box the screen some how so the processor isn't having to draw a full screen, this will have no effect on battery.
In fact, assuming the GPU does process the image at 1080 first, it is having to process it again to make it fit the 1440 screen. Stands to reason that would cost you battery life.

mr pnut said:
I feel you..... but serious tho you can't see no difference at all.
Click to expand...
Click to collapse
But serious though...the difference in battery life is most-likely extremely negligible.
The display is still pushing power to all of the pixels...the only battery you're saving is the CPU working slightly harder to create higher-resolution outputs. This is about the same as changing the CPU frequency to like 5% less than normal...you might gain another ~15 minutes of screen on-time.

I had the wonderful sensor Ind wake lock today, so battery life definitely sucked. I'm hitting 5 hours SOT with 15% left as I type this. That's about 1.5 hours SOT shy of an average day with no wake lock issues and 1440p. As we all knew battery life would get at best a very minor bump with this, the screen is still lighting all the pixels after all. Oh well, it was worth trying out, but if Google would just fix these dam wake locks it would save many of us a ton of battery life! Back to beautiful 1440p