Hello everyone,
This thread will be used as a hub where I share some discoveries/observations which I stumbled upon mostly during working on my kernel projects.
I´ll clone the same thread over to the Pixel 7 forum as well. So without much further ado let´s just dive right into it.
A year ago everyone was excited for the Google SoC called Tensor 1 called GS101. One year later there is Tensor 2 called GS201.
I suggest to read about the differences, updated modem, ISP, TPU and GPU in various tech related articles.
Here´s a table so everyone gets up to speed on cores used, max freqs and other details:
{
"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"
}
But how does that translate on the devices?
There were quite a few rumors before the actual release of the Tensor 2 SoC being manufactured on 4nm Samsung node instead of 5nm. However that was just wild speculation and unfortunately turned out to be not accurate. Tensor G2 is still manufactured in the 5nm process as confirmed by Google. This was quite a negative surprise to myself, as I don´t have good experiences from SD888 that´s also being manufactured in Samsung 5nm node and is quite a hot chip. While the switch to Samsung 4nm node, wasn´t all that great either (check sd8 gen1 on samsung 4nm vs sd8+ gen1 on tsmc 4nm) it would still have been an improvement.
While I was very excited for the Tensor G1 when the Pixel 6 devices launched, that excitement ebbed down the work I worked on the Pixel 6 series. The more I learned about the source, the more I stumbled upon Exynos driver over exynos driver, some are just left exactly like on Exynos device, some were "re-branded" by Google. Some Google did customize, but most of the drivers are just very much Exynos.
So all in all the following excerpt from Andrei Frumusanu´s article here sums it up pretty fitting:
While Google doesn’t appear to want to talk about it, the chip very clearly has provenance as a collaboration between Google and Samsung, and has a large amount of its roots in Samsung Exynos SoC architectures.
Click to expand...
Click to collapse
The same is true for the Tensor 2, despite minor upgrades there. As I learned over the time, Tensor shares a lot of Exynos characteristics, one of those is performance vs thermals as hinted by in the linked Anandtech article. So let´s just jump into that first topic I want to cover.
I will cover more topics, those will also be probably interconnected to each other, but we have to start somewhere.
Thermals, Thermal Ceiling, Exynos Roots and Maximal Performance
To start things of: Thermals is a term that actually sums up a few mechanisms. Lets split this into two main areas.
The thermal ceiling (let´s call it that) that´s being implemented in the kernel, as the maximal temperature the SoC is allowed to be operated at.
The thermal-hal uses combined sensors, also virtual sensor, and restricts different subsystems, based on the temperature of those sensor. Those can be called skin temps, shell-temps, battery temperature, modem temperature etc.
First let´s explore the thermal ceiling on the two SoCs:
GS101 on Pixel 6 devices is allowed to be operated at 90°C. GS 201 on Pixel 7 devices raised the thermal ceiling by 10°C to 100°C.
If changes to the internal design allowed them to raise this, without further increasing heatup of the device, or if they just applied changes to the thermal-hal to better keep this in check I don´t know at this moment.
Let´s get back to the Exynos characterstics. I talked to a few other developers I met along that way with Exynos "experience". Exynos SoCs reach the thermal ceiling extremely quickly, as I learned. This means, the SoC can´t keep its max CPU freqs for more than a few seconds without touching the thermal ceiling and getting restricted. This is in a way also the case for other SoCs, but Exynos is very extreme in this regard. But it´s just the characteristics of the SoC, like previously mentioned.
That means in turn: The thermal ceiling is setting the maximal performance allowed, to a great extent. If the thermal ceiling is raised, the maximum performance can be held longer.
Here´s a demonstration of this:
Pixel 6 Pro in its default configuration running at 90°C temp ceiling:
New video by freak 07
photos.app.goo.gl
Pixel 6 Pro with temperature ceiling raised to 100°C, instead of 90°C running at Pixel 7 Pro clocks
New video by freak 07
photos.app.goo.gl
Pixel 7 Pro, with the default configuration of 100°C
New video by freak 07
photos.app.goo.gl
Now what´s interesting, the big cores get the hottest at the quickest rate. Once the ceiling is reached, the max performance drops, as maximal performance will be restricted by restrict max cpu freqs.
That´s the case after a few seconds on both SoCs, in typcial Exynos fashion.
Let´s make the next connection:
Although I´m not necessarily a friend of benchmark apps, how does that change the results of a CPU oriented benchmark like Geekbench you might ask yourself. There are other benchmarks, but I want to keep this simple for now.
The answer is: The Pixel 6 Pro with GS101 gets pretty close to the results of the Pixel 7 Pro with GS201.
So for comparisons sake:
On the left Pixel 6 Pro in its default configuration running at 90°C temp ceiling.
On the right Pixel 6 Pro with temperature ceiling raised to 100°C, instead of 90°C running at Pixel 7 Pro clocks.
The kernel used was the same, no changes to anything that could impact performance.
The left screenshot above was taken from the Pixel 7 Pro review from XDA, while the right one was taken on my Pixel 7 Pro running my kernel.
Please don´t start benchmark contests now, It´s just for comparisons sake.
It makes sense for single-core to be less impacted, as single core benchmarks don´t put as much thermal pressure on the SoC -> not touching the thermal ceiling as much and therefore no cutback are applied.
Geekbench applies a series of short benchmarks to the device. Usually not longer than 3-8 seconds, which is ideal for a SoC like the Tensor. Short bursts with max performance, so it can run "nearly" without touching the thermal ceiling.
If a benchmark is structured differently, like the CPU stress test you will see QCOM SoCs holding their max-freqs for minutes, instead of seconds.
Well that´s the first part. More to come. I hope everyone enjoyed this little writeup so far.
I wish everyone a nice evening.
Thermal Ceiling/Maximal Performance - A comparison between QCOM Snapdragon and Tensor
Now you might ask yourself, how does QCOM´s Snapdragon behave in the little test we conducted above.
You can find the answer below.
For this a Zenfone 9 with the Snapdragon 8 + Gen 1 is used.
Pixel 6 Pro in its default configuration running at 90°C temp ceiling:
New video by freak 07
photos.app.goo.gl
Pixel 7 Pro, with the default configuration of 100°C:
New video by freak 07
photos.app.goo.gl
Asus Zenfone 9, with the default configuration of 110°C temp ceiling:
New video by freak 07
photos.app.goo.gl
As you can see the Zenfone 9 with SD 8+ Gen 1 can hold the max-freqs for minutes. It doesn´t touch the thermal ceiling when running under max load for a minute, while Tensor immediately scratches the ceiling.
I´m not a SoC expert and I think only engineers with insider knowledge know the exact reason why Exynos based SoCs behave that way. They just seem to work totally different in that regard.
Another point is, since the SoCs are different we can´t compare the temperatures one to one. There´s no way for me to know the exact placement of the temperature sensors, all I can say for sure is the SD 8+ Gen 1, does not touch the thermal ceiling in this test and there seems to be a lot of headway after one minute of maxed out CPU.
In the end the result will be the same. After a while the device will heat up and the thermal-hal will throttle the ZF9 back as well, as with only passive cooling that´s inevitable.
this one is mine too
this one too
number 5
and the last one
Interesting, thanks for this explanation and comparison. Learned something new today.
so tensor is just Exynos but rebranded and more ai performance
w_tapper said:
so tensor is just Exynos but rebranded and more ai performance
Click to expand...
Click to collapse
I´m not sure if it´s that easy in the end. It´s not just simply rebranded, there´s a lot of custom stuff in it, but at the core it´s based on the Exynos that´s clear.
I guess it´s just not possible to come up with a real custom chip in only a few years. It´s probably the beginning of that.
A few thoughts on tensor, that already allow google to better tune software to hardware which results in real life benefits:
The 4+2+2 design. That´s unique to the tensor and greatly favors those short performance boosts that are really critical in everyday usage. Everyone is constantly opening apps, if those apps launch a fraction faster, that´s a good real world benefit. (typical app launch times are between 0.2s and 0.7s.)
Tensor is not lacking behind the competition, let´s take the SD8+ Gen1, at all in that regard as just looking at benchmark scores would suggest.
I guess that´s a good section for another in depth post, that investigates real world usage.
The other is the TPU and the ISP.
Last year I used a Sony Xperia 1 III with SD888 (Sony more or less uses very relaxed thermal and basically unleashes full perf of SD888) vs the Pixel 6 Pro to edit the same video file and cut it.
The Pixel 6 Pro finished the task quite a bit faster. If I recall around 10-15 second faster.
There are other examples, but this will do for now.
updated the second post, with a comparison between QCOM´s Snapdragon and Google´s Tensor
Interesting since I wonder if a switch from P6P to P7P is worth it.
Utini said:
Interesting since I wonder if a switch from P6P to P7P is worth it.
Click to expand...
Click to collapse
I like my P7P even better than the P6P, although I had no particular issues with the P6P. The USB 3.2 Gen 2 versus Gen 1 USB-C port pushed me over the edge.
Utini said:
Interesting since I wonder if a switch from P6P to P7P is worth it.
Click to expand...
Click to collapse
In the end you have to decide yourself.
I can give maybe a few experiences. I think there´s not a single real regression going from p6pro to p7pro for me. That´s an important point, considering that´s not always the case in such upgrade scenarios.
I like the slightly lower placement of the buttons, they´re easier to reach for me. The FP might be a tad faster, but I was one of the lucky ones that never really had trouble on the p6pro. I like the face unlock.
What I really enjoy is the macro mode in GCAM. It works quite well and fills a gap.
While I never had any real problems with mobile network on the p6pro there are three areas I regularly drive through. All phones struggle there in a way. Some lose signal there, amongst them the p6pro, other can barely keep it. Though you never know with latency and "cheating" with the signal icons. Each OEM handles that differently.
However usually I notice I can´t call anyone there and spotify playback gets interrupted due to signal loss on p6pro. With the p7pro I´m able to make and receive calls in those areas and also no other kind of signal loss. If that´s the case for everyone or just a border case for me I don´t know. It´s just an observation.
There are lots of small improvements that add up. Is it worth it for you, you have to decide.
Freak07 said:
In the end you have to decide yourself.
I can give maybe a few experiences. I think there´s not a single real regression going from p6pro to p7pro for me. That´s an important point, considering that´s not always the case in such upgrade scenarios.
I like the slightly lower placement of the buttons, they´re easier to reach for me. The FP might be a tad faster, but I was one of the lucky ones that never really had trouble on the p6pro. I like the face unlock.
What I really enjoy is the macro mode in GCAM. It works quite well and fills a gap.
While I never had any real problems with mobile network on the p6pro there are three areas I regularly drive through. All phones struggle there in a way. Some lose signal there, amongst them the p6pro, other can barely keep it. Though you never know with latency and "cheating" with the signal icons. Each OEM handles that differently.
However usually I notice I can´t call anyone there and spotify playback gets interrupted due to signal loss on p6pro. With the p7pro I´m able to make and receive calls in those areas and also no other kind of signal loss. If that´s the case for everyone or just a border case for me I don´t know. It´s just an observation.
There are lots of small improvements that add up. Is it worth it for you, you have to decide.
Click to expand...
Click to collapse
Ye I have read about the new "features". I loved face unlock on the 4XL and all the other new changes on the P7P are very welcome as well. But those aren't enough for me atm to switch to a new device. Especially since I first want thermals (e.g. 4k 60 fps video recording or while using the phone in the summer), radio signal and battery to be improved.
Radio signal seems to be better for now. My problems with the P6P seem to be the same that you have/had
Maybe that also helps slighty with battery, especially in dual-sim mode.
But I guess ultimately I will wait for the P7P and hope that it is going to be a "bigger" jump. The P6P is really fine for a phone. Jesus, I would even be fine with the P4XL.
Nice research and information. Was very interested in taking a deeper dive into the core differences between the G1 and G2. Thanks for sharing and laying this all out in a way that makes sense. Interested to see what else you can uncover!
Wait ... how did you change the temperature control threshold? I have been working on this. After the surface temperature of the phone exceeds 39 ° C, the operating frequency of Tensor will drop sharply, making the game experience very bad ... I think there is something like Magisk module to increase temperature control? Translate form Google.
Related
Hey guys this is Saad from Windows Phone Daily,
I thought I'd share this on here since it doesn't look like any of the bigger WP7 sites have picked up on this yet.
A Nokia 900 "Hydra" just leaked from Daily Mobile Forum and it has some really impressive specs: a 4" CBD AMOLED, 1.5 GHz dual-core(!) processor, 720p display, 12MP Carl Zeiss camera, 1080p video recording, and new "swipe gestures" in the OS (maybe something like Windows 8).
More details can be found in our article: http://www.windowsphonedaily.com/2011/10/rumor-nokia-900-hydra-packs-dual-core.html
aaaand here's the original source: http://forum.dailymobile.se/index.php?topic=52961.0
Enjoy!
{
"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"
}
too good to be true, but hey
The fact that it claims to have "720p screen resolution" is proof enough that this is not real.
Well there's fake, and then there's this.
First of all the resolution. 720p screen resolution would be a direct violation against all of Microsoft's chassis specifications. The resolution is limited to 800x480 for a reason, to make it easier for developers. If this were true Microsoft would be on the very slippery slope of fragmentation.
To clarify: The platform is currently designed for one resolution and that resolution only. More or less the only possible way to continue from this is screens with the resolution 1600x960, exactly doubled on both sides. This was the way that Apple upgraded the screen between iPhone 3G(S) and iPhone 4 and it was for a very good reason: To keep app compability. If Microsoft were to allow a new resolution almost all apps would need a patch for that new resolution. Would Microsoft kill a majority of their Marketplace? no.
None of the current chassis specifications define any support for a Dual Core processor, so that is not likely at this time, maybe next update.
"Dedicated graphic adapter" sounds a bit fishy. It sounds to a reference to a dedicated GPU chip as opposed to todays SoC chips where everything is integrated on one chip. I'm not an expert in the area but it certainly sounds strange that someone suddenly would design a chip very differently from everyone else.
4 Sterophonic loudspeakers. I think that this certainly sounds strange, not sure what they mean with that.
USB 3.0 and NFC doesn't feel very likely. On the other hand Nokia is able to do this with the special privileges Microsoft gave them. I'm not as certain on this one as the others.
Also, don't you all think that the picture doesn't look very professionally done?
Source post
This was up almost two months ago and a whole load of unresponsible news sites did write the news like this was a fact, absolutely disregarding how obviously fake it is.
Why wont apps be compatible with different resolutions? On Android they work well regardless of what resolution. Whats different in WP?
Sir. Haxalot said:
First of all the resolution. 720p screen resolution would be a direct violation against all of Microsoft's chassis specifications. The resolution is limited to 800x480 for a reason, to make it easier for developers. If this were true Microsoft would be on the very slippery slope of fragmentation.
To clarify: The platform is currently designed for one resolution and that resolution only. More or less the only possible way to continue from this is screens with the resolution 1600x960, exactly doubled on both sides. This was the way that Apple upgraded the screen between iPhone 3G(S) and iPhone 4 and it was for a very good reason: To keep app compability. If Microsoft were to allow a new resolution almost all apps would need a patch for that new resolution. Would Microsoft kill a majority of their Marketplace? no.
Click to expand...
Click to collapse
Actually, 1200x720 would be a 1.5X increase, so this would work too. The OS could take care of resizing/resampling of graphics resources, while text rendering would automatically be improved. Devs could also upgrade their apps by adding higher-res resources, but it wouldn't break compatibility with older phones. Apple needed to double things because 3GS and older phones had incredibly crappy low-res screens.
I'm sure this will happen anyway, but I doubt that's what MS and OEMs have in mind until the next big update, and I don't think it'll happen in Q1.
On the contrary, judging by the leaks, the next update will be all about decreasing price.
Interesting how the back and search buttons are missing too. Also, the technical specification seems to include a price.
Looks like the picture was drawn by a kid
If true this may be my next device
Almost no bezel at the corners? No way, that looks fake as hell. The only specs I'm really interested now are expendable memory and screen resolution. And a bit more RAM for extended multi-tasking, maybe.
I can't believe people are actually conversing about this trash.
I also can't believe I typed this reply.
And screen resolution doesn't matter in WP7 as the OS uses a scalar. Things would be automatically upscaled. It's why the phone's UI performs well on the launch specs. The resources the phone displays is actually lower resolution that you'd assume they are. iOS doesn't use a scalar so resources have to be appropriately scaled for the display resolution. That's why developers needed to update for iOS5 and develop new iPad apps. Android is similar.
N8ter said:
I can't believe people are actually conversing about this trash.
I also can't believe I typed this reply.
And screen resolution doesn't matter in WP7 as the OS uses a scalar. Things would be automatically upscaled. It's why the phone's UI performs well on the launch specs. The resources the phone displays is actually lower resolution that you'd assume they are. iOS doesn't use a scalar so resources have to be appropriately scaled for the display resolution. That's why developers needed to update for iOS5 and develop new iPad apps. Android is similar.
Click to expand...
Click to collapse
I don't know what you are talking about. I've looked into a Dumped ROM & extracted the PNG's from the DLL's & they are full size. They are NOT low res images scaled up. If they were, they'd look blurry like early WVGA Titanium Themes that were just edited WQVGA CPR's with no changes to the graphics. The Tiles & Icons are 173x173/62 x 62 pixels. That is full size. The 1st Part Tiles & Icons are a bit smaller only because they are just the Icons inside the Tiles, not the whole Tile like for 3rd Party apps. Developers have been making their Tiles cheaply. You'll notice the Tile image is called Background, meaning it's the background of the Tile Image & there should be a separate Tile Icon that goes "over" it.
To get back on topic, I highly doubt this is real as per resolution unless MS actually gave Nokia special permissions to edit the UIB/UIX files. They are pretty much like the CPR's of Titanium, just compiled so we can't edit them ourselves yet. We have no tools to decompile/recompile them.
Wouldn't be as unbelievable if this was a Q3 device (though, ideally we'll have faster by then), but a Q1? Naaaah
Nearly every spec isn't supported. Fake.
Sir. Haxalot said:
First of all the resolution. 720p screen resolution would be a direct violation against all of Microsoft's chassis specifications. The resolution is limited to 800x480 for a reason, to make it easier for developers. If this were true Microsoft would be on the very slippery slope of fragmentation.
To clarify: The platform is currently designed for one resolution and that resolution only. More or less the only possible way to continue from this is screens with the resolution 1600x960, exactly doubled on both sides. This was the way that Apple upgraded the screen between iPhone 3G(S) and iPhone 4 and it was for a very good reason: To keep app compability. If Microsoft were to allow a new resolution almost all apps would need a patch for that new resolution. Would Microsoft kill a majority of their Marketplace? no.
None of the current chassis specifications define any support for a Dual Core processor, so that is not likely at this time, maybe next update.
"Dedicated graphic adapter" sounds a bit fishy. It sounds to a reference to a dedicated GPU chip as opposed to todays SoC chips where everything is integrated on one chip. I'm not an expert in the area but it certainly sounds strange that someone suddenly would design a chip very differently from everyone else.
4 Sterophonic loudspeakers. I think that this certainly sounds strange, not sure what they mean with that.
USB 3.0 and NFC doesn't feel very likely. On the other hand Nokia is able to do this with the special privileges Microsoft gave them. I'm not as certain on this one as the others.
Also, don't you all think that the picture doesn't look very professionally done?
Source post
This was up almost two months ago and a whole load of unresponsible news sites did write the news like this was a fact, absolutely disregarding how obviously fake it is.
Click to expand...
Click to collapse
Wow you're really out to get me Sir. Haxalot!
I am aware that this is from August, and I apologize for reporting on it as though it was new. However, this is not "obviously fake". Like I said in the article and my reply to you in the comments section, it is very likely that these specs will be introduced in the future. There's no reason why you can rule that out, unless you're Microsoft's lead Windows Phone manager.
It is a rumor and yes there are some things that do stretch the limits of reality--mostly the Q1 release which is completely ridiculous. BUT Nokia hasn't denied it and Microsoft hasn't denied it, so until then I will report on it.
OGCF said:
Wow you're really out to get me Sir. Haxalot!
I am aware that this is from August, and I apologize for reporting on it as though it was new. However, this is not "obviously fake". Like I said in the article and my reply to you in the comments section, it is very likely that these specs will be introduced in the future. There's no reason why you can rule that out, unless you're Microsoft's lead Windows Phone manager.
It is a rumor and yes there are some things that do stretch the limits of reality--mostly the Q1 release which is completely ridiculous. BUT Nokia hasn't denied it and Microsoft hasn't denied it, so until then I will report on it.
Click to expand...
Click to collapse
the problem is you're "reporting" something that isn't real by any stretch of the imagination. fake nonsense rumors like this fly so far below the radar of companies like MSFT and Nokia that they don't even know it exists, much less have the desire to deny it.
looks like kids drawing or someone in the night market drawed it...lol
This is fake. I thought everyone knew that. It's embarrassing that you would post it on your website EVEN with the disclaimer that it's just a rumor. This mockup is brought to you by the same fool that opened up paint for this one too:
Sir. Haxalot said:
First of all the resolution. 720p screen resolution would be a direct violation against all of Microsoft's chassis specifications. The resolution is limited to 800x480 for a reason, to make it easier for developers. If this were true Microsoft would be on the very slippery slope of fragmentation.
To clarify: The platform is currently designed for one resolution and that resolution only. More or less the only possible way to continue from this is screens with the resolution 1600x960, exactly doubled on both sides. This was the way that Apple upgraded the screen between iPhone 3G(S) and iPhone 4 and it was for a very good reason: To keep app compability. If Microsoft were to allow a new resolution almost all apps would need a patch for that new resolution. Would Microsoft kill a majority of their Marketplace? no.
None of the current chassis specifications define any support for a Dual Core processor, so that is not likely at this time, maybe next update.
"Dedicated graphic adapter" sounds a bit fishy. It sounds to a reference to a dedicated GPU chip as opposed to todays SoC chips where everything is integrated on one chip. I'm not an expert in the area but it certainly sounds strange that someone suddenly would design a chip very differently from everyone else.
4 Sterophonic loudspeakers. I think that this certainly sounds strange, not sure what they mean with that.
USB 3.0 and NFC doesn't feel very likely. On the other hand Nokia is able to do this with the special privileges Microsoft gave them. I'm not as certain on this one as the others.
Also, don't you all think that the picture doesn't look very professionally done?
Source post
This was up almost two months ago and a whole load of unresponsible news sites did write the news like this was a fact, absolutely disregarding how obviously fake it is.
Click to expand...
Click to collapse
Right on the money here. 100% fake.
Yes it's quite obvious the photo and what not is fake. However, someone email this to Nokia and make it happen for Apollo.
So...., are we getting this 8 core update or what?
Doesn't seem like it , it's been confirmed that note 3 will not be getting the 8 core patch...since it's in essence a smaller version of the note 10.1 , well you do the math.
Can someone prove me wrong?? (please)
There are both thermal and battery life concerns. If Samsung thought this would up the anti for performance and not compromise stability or battery life, they would probably release it.
it would be nice to turn off the weaker cores then. i'd suspect them to be a cause of a lot of the lagginess and frame drops in video players. bs player has been the smoothest so far.
I believe there were a few articles floating around specifically saying that the CPU in the note 10.1 2014 edition could potentially get the true octacore mode. I believe there were also demo videos using this tablet in another thread.
ChrisNee1988 said:
I believe there were a few articles floating around specifically saying that the CPU in the note 10.1 2014 edition could potentially get the true octacore mode. I believe there were also demo videos using this tablet in another thread.
Click to expand...
Click to collapse
Here's the thread from the N3 forum showing a tablet running HMP. It's clearly a test mule and not a N10.1-14. Here's the bottom line about adding HMP after the fact - there's nothing in it for Samsung. All the current products are marketed with a defined "high end" performance capability and are vaguely marketed as "four core" for S-800 and "1.9+1.3" for Octa. What's the benefit to Samsung of lighting up all eight cores to exceed today's performance and to take on the burden of the impact to battery life and potential thermal issues? All for something that only enthusiasts (us) know or care about.
The only clear benefit of implementing HMP is adding the four A7 cores on top of the four A15 cores to improve high-end performance which will end up tanking today's battery life. As implemented, 1-4 cores of each cluster can run within a given cluster and the gains of mixing and matching clusters (EG: 4 A7+1 A15 vs. 2/3 A15) is unproven and questionable.
So maybe future Octa chips will run HMP but, primarily because there's no benefit to Samsung of doing anything to existing devices, I don't see it being made retroactive.
http://forum.xda-developers.com/showthread.php?t=2556264
madsquabbles said:
frame drops in video players. bs player has been the smoothest so far.
Click to expand...
Click to collapse
With what kind of videos do you experience frame drops? I use my tablet very often while using my cross-trainer to watch video and never experienced any frame drops. I tried MX Player, Dice Player and VPlayer and they are all smooth - actually no wonder, because they all use the HW acceleration for playback which can handle HD content without any problems.
aac and ac3 seem to cause the most problems SD and HD. I generally keep the nitrate around 2.5 Mb/s just for portability sake. container doesn't matter as avi, mkv, and mp4 all have the drops. I've used mx, BS, Archos, xbmc, and a few others and installed custom codecs if needed. heck even Netflix has a few frame drops. maybe I'm just too sensitive to it, but my s4, gn2, and gn8 have no problems. my old gn10.1 had no problems either. this 10.1 2014 is also the first Samsung device I've had that doesn't natively support ac3. hardware playback with alternative players do play the ac3 audio, but with dropped frames.
So does the note 2014 doesn't have HMP?
Does the new one 12.2 has it.
madsquabbles said:
aac and ac3 seem to cause the most problems SD and HD.
Click to expand...
Click to collapse
That's strange, almost all my video files contain ac3 6-channels audio and playback is smooth without any frame drops.
DeBoX said:
Doesn't seem like it , it's been confirmed that note 3 will not be getting the 8 core patch...since it's in essence a smaller version of the note 10.1 , well you do the math.
Can someone prove me wrong?? (please)
Click to expand...
Click to collapse
I believe there was a demo at CES with hmp running on the Note 10.1. I'm almost positive. Where was it confirmed the note 3 was not getting it? Just curious.
havekk said:
I believe there was a demo at CES with hmp running on the Note 10.1. I'm almost positive. Where was it confirmed the note 3 was not getting it? Just curious.
Click to expand...
Click to collapse
If Samsung was going to add HMP to the current generation of Exynos they would have tied it to something big, like the launch of a new $750+ tablet. HMP was shown on a test tablet, not the N10.1-14. Idealists keep hoping but it's not likely to happen.
BarryH_GEG said:
If Samsung was going to add HMP to the current generation of Exynos they would have tied it to something big, like the launch of a new $750+ tablet. HMP was shown on a test tablet, not the N10.1-14. Idealists keep hoping but it's not likely to happen.
Click to expand...
Click to collapse
Do you happen to know what SoC the test tablet was running? Everything I read said it was the 5420. Btw, I don't own a Note 10.1 - I still have a ****ty iPad 3. I haven't heard anything from "idealists" just going over what I have read from folks at the event. From what I have read, it seems as likely to happen on the Note 10.1 and Note Pro as it is to NOT happen. I have heard it won't happen on the Note 3; however, I assume that is due to size and heat issues. Shouldn't be the case with the lager tablets.
Only time will tell. You or I certainly don't know. I'm gonna wait to buy a new tablet and keep an eye on it though.
Thanks,
havekk said:
Do you happen to know what SoC the test tablet was running? Everything I read said it was the 5420. Btw, I don't own a Note 10.1 - I still have a ****ty iPad 3. I haven't heard anything from "idealists" just going over what I have read from folks at the event. From what I have read, it seems as likely to happen on the Note 10.1 and Note Pro as it is to NOT happen. I have heard it won't happen on the Note 3; however, I assume that is due to size and heat issues. Shouldn't be the case with the lager tablets.
Only time will tell. You or I certainly don't know. I'm gonna wait to buy a new tablet and keep an eye on it though.
Thanks,
Click to expand...
Click to collapse
There's not one commercial/business reason for Samsung to update tablets already on the market beyond their published specs. Zip. Zero. Nada. And there's a reason not to. If an update to the SoC goes wrong and borks the tablet they have the liability of repairs. Enthusiasts (us) know what HMP is. The masses don't know, don't care, and are fine with what they're getting/got.
BarryH_GEG said:
There's not one commercial/business reason for Samsung to update tablets already on the market beyond their published specs. Zip. Zero. Nada. And there's a reason not to. If an update to the SoC goes wrong and borks the tablet they have the liability of repairs. Enthusiasts (us) know what HMP is. The masses don't know, don't care, and are fine with what they're getting/got.
Click to expand...
Click to collapse
Ok. I can see you're really passionate about this not happening lol. So for the sake of not getting another response, I'll concede that you are absolutely definitely correct about this. It is not going to happen.
Thanks for you assistance on this matter.
havekk said:
Ok. I can see you're really passionate about this not happening lol. So for the sake of not getting another response, I'll concede that you are absolutely definitely correct about this. It is not going to happen.
Thanks for you assistance on this matter.
Click to expand...
Click to collapse
The N10.1-14's battery life is "adequate" and the Exynos version takes forever to charge. The only benefit of HMP is adding the A7 cores on top of the A15 cores at max load. That's going to take a toll on battery life and increase the thermal load beyond what the device is designed for possibly impacting component life. Forgive me if I'm not enthused.
Ok, I'm out.
BarryH_GEG said:
The N10.1-14's battery life is "adequate" and the Exynos version takes forever to charge. The only benefit of HMP is adding the A7 cores on top of the A15 cores at max load. That's going to take a toll on battery life and increase the thermal load beyond what the device is designed for possibly impacting component life. Forgive me if I'm not enthused.
Click to expand...
Click to collapse
That's interesting to know. All of the test I have read about prior had gains in battery life by using the cores much more efficiently. You posted a video describing this process in detail and how it helped battery life, then you post the exact opposite opinion?
Something else I find interesting is that you say implementing HMP would "increase the thermal load beyond what the device is designed for" Yet... we all know that HMP is hardware-enabled in the 5420 SoC. So couldn't it be said that the device was designed to use HMP? I think it could.. nay, it should lol.
Wait a sec! I get it!! I just saw that you have fought this battle before with Iba21 - You really seem to have something against this whole thing lol. From what I just read, Iba21 really pooped on your entire argument and you stopped responding.
Don't worry about responding as I'm getting out of this "discussion". It's clear you have an opinion as do I.. Only time will tell who's is more accurate.
havekk said:
That's interesting to know. All of the test I have read about prior had gains in battery life by using the cores much more efficiently. You posted a video describing this process in detail and how it helped battery life, then you post the exact opposite opinion?
Something else I find interesting is that you say implementing HMP would "increase the thermal load beyond what the device is designed for" Yet... we all know that HMP is hardware-enabled in the 5420 SoC. So couldn't it be said that the device was designed to use HMP? I think it could.. nay, it should lol.
Wait a sec! I get it!! I just saw that you have fought this battle before with Iba21 - You really seem to have something against this whole thing lol. From what I just read, Iba21 really pooped on your entire argument and you stopped responding.
Don't worry about responding as I'm getting out of this "discussion". It's clear you have an opinion as do I.. Only time will tell who's is more accurate.
Click to expand...
Click to collapse
Saying inflammatory things and then posting "no need to reply" is passive aggressive. There are two possible benefits of HMP - 1) using combinations of A7 and A15 cores across clusters and 2) going beyond the current max of 4 A15 cores. Do you know if there's battery life to be saved by using 2 A7 cores and 1 A15 core as opposed to using 2 A15 cores? I don't but I'm guessing the advantage to be minimal. You don't have to be an engineer to understand that using cores in addition to the current max of 4 A15's is going to create more heat and draw more power.
Just a thought, but allowing HMP during periods of high load and having all 8 cores online all the time represent two different scenarios. If core OS functions could remain on an A7 and yield all 4 A15s to applications, you could see an improvement purely due to a reduction in OS interrupt servicing. The case for thermal dissipation and battery life represents an unknown as to how much of an impact it would really have. It would be highly dependent upon how heavily you are using the active cores. The penalty would be proportionate to how hard you're pushing it.
All that said, I tend to agree with BarryH_GEG. Samsung hasn't demonstrated a lot in the forward thinking department. They make capable tablets, but just don't have it together on a lot of fronts and definitely don't strive to optimize products in a way that delivers maximum performance. Their focus is primarily on gimmicks and visual features and hitting a point of "acceptable", not exceptional performance. I'm not bashing, I have owned numerous galaxy line products and will continue to do so until a viable active digitizer tablet surpasses the note line. It's just a business play on their part. You expend enough development resources to get a mainstream sale, no more.
The note 12.2 is a prime example. At the time of release, Samsung has no book cover available, poor planning. They didn't have Hancom Office in the preloaded image and it wasn't available until a day after release, also poor planning. The end result is a customer experience that isn't smooth and demonstrative of a company that gives significant consideration to first impressions or user experience.
I have had my Note 2014 (32GB) for two days and would not want to have both core sets running at the same time. The device can already get very warm in the SOC area and the battery drains just like my iPad 4 when playing 3D games- faster than I want it to. Games are smooth and ditto for apps, so not seeing an advantage of all eight, but see the two key disadvantages. Not running into any core hand-off issues some are reporting that can cause lag. Not yet anyway.
Added:
I use Nova for my launcher, since IMO much better than Samsung's. Based on performance, I see no reason for me to root the 2014. Very happy already, so see no need to rock the boat.
Greetings xda
I just purchased a secondhand/used/pre-owned Verizon model LG G Pad 8.3 (vk810) on ebay and have noticed that sinnce day one that I've had it (maybe 4 days?), that overall performance is pretty inconsistent. Sometimes it flies through the UI with impressive ease and other times it seems bogged down simply running Chrome with a single tab open and hitting the Home button.
First thing I did when it arrived was run some benchmarks. I noticed that across the board I was receiving results that were roughly 40-50% lower than all the charts I've seeen others posting online, including proffessional reviewers' articles.
I cannot recall *all* the scores for each of the different programs, but for instance in Antutu it seems everyone else is hitting 25000+, whereas my device is sitting at around 15000 or so. I did this on stock 4.2.2, 4.4.2, rooted, unrooted, fresh stock rom flash, with TWRP and without, and each time it's been the same outcome. Antutu reports all the same hardware everyone else's does, yet this. I'm fairly certain it wasn't some kind of bait and switch for a knockoff or inferior tablet.
When gaming, most of the stuff I play runs decent enough, with some stutter here and there, albiet they are not all demanding games (from Hay Day, to Walking Dead, to CoD: Strike Team and everything in between), but I've still been noticing more overall chug factor than the specs of this tablet should be delivering all day with ease. I started to worry something was wrong when I was strugging to run GTA: San Andreas on Medium. When I look at say the Antutu example again, other people's G Pad 8.3 rank shy of a Galaxy S4, where on mine I'.m getting beat our by the Galaxy S III.
The tablet doesn't appear to be overheating, causing throttling; but I just can't seem to wrap my head around it or even know where to begin troubleshooting other than coming here and asking for assistance. If someone could possibly be so kind enough, any thoughts on how I should proceed in troubleshooting this? Thanks in advance!
Kaiyo Droid said:
First thing I did when it arrived was run some benchmarks. I noticed that across the board I was receiving results that were roughly 40-50% lower than all the charts I've seeen others posting online, including proffessional reviewers' articles.
I cannot recall *all* the scores for each of the different programs, but for instance in Antutu it seems everyone else is hitting 25000+, whereas my device is sitting at around 15000 or so. I did this on stock 4.2.2, 4.4.2, rooted, unrooted, fresh stock rom flash, with TWRP and without, and each time it's been the same outcome. Antutu reports all the same hardware everyone else's does, yet this. I'm fairly certain it wasn't some kind of bait and switch for a knockoff or inferior tablet.!
Click to expand...
Click to collapse
The Verizon 8.3/VK810 is only clocked at 1.5GHz vs the 1.7GHz of the V500/510.
Planterz said:
The Verizon 8.3/VK810 is only clocked at 1.5GHz vs the 1.7GHz of the V500/510.
Click to expand...
Click to collapse
Interesting... Thanks for the info, Planterz Wasn't aware of that.
I'm certainly no expert when it comes to CPU's, especially in the mobile world. Would a mere 200Mhz per core difference in clock speed provide such disparaging results among otherwise identical hardware?
Kaiyo Droid said:
Interesting... Thanks for the info, Planterz Wasn't aware of that.
I'm certainly no expert when it comes to CPU's, especially in the mobile world. Would a mere 200Mhz per core difference in clock speed provide such disparaging results among otherwise identical hardware?
Click to expand...
Click to collapse
15 000 antutu is in range of s3, I think, which has inferior hardware! You should have scored easily over 21-22000 with this beast's snapdragon 600 no matter even with a 200mhz difference! Charge the battery to full, switch off for about 2 hours, boot, standby for about 1 hour and check again! If still under 20 000, something is wrong with device and you should change it!!
Fullmetal Jun said:
15 000 antutu is in range of s3, I think, which has inferior hardware! You should have scored easily over 21-22000 with this beast's snapdragon 600 no matter even with a 200mhz difference! Charge the battery to full, switch off for about 2 hours, boot, standby for about 1 hour and check again! If still under 20 000, something is wrong with device and you should change it!!
Click to expand...
Click to collapse
I only get around 20000 on my V500, but I'm still running stock (rooted) 4.2.2. I am about to finally updated to 4.4 and flash the updated Dalvik and Qualcomm binaries though.
I don't know why Verizon's version is only clocked at 1.5GHz, when the SD600 is clearly capable of much more (the Galaxy S4 has the SD600 clocked to 1.9GHz). Maybe they saved a few bucks by going with a lower bin, or maybe it's just underclocked to stretch out the battery life. 200MHz across 4 cores definitely makes a difference in benchmarks, especially with the amount of pixels they have to push on the 8.3 (1920x1200 makes it 1200p, not the standard 1080p of FHD). The Galaxy S3 might get 15000 in Antutu, but it also has less than half the pixels to push. My Galaxy Light, which is NOT a high performance device, also gets 15000-16000, but it has less than half the pixels of the S3. Remember, benchmarks aren't always an indicator of real-world performance.
EDIT: I haven't used the Verizon 8.3, but being a Verizon device, I imagine it's chocked full of the typical Verizon bloat. Check your processes to see if there's something unnecessary running in the background and delete that crap, or at least disable or Greenify it.
Planterz said:
I only get around 20000 on my V500, but I'm still running stock (rooted) 4.2.2. I am about to finally updated to 4.4 and flash the updated Dalvik and Qualcomm binaries though.
I don't know why Verizon's version is only clocked at 1.5GHz, when the SD600 is clearly capable of much more (the Galaxy S4 has the SD600 clocked to 1.9GHz). Maybe they saved a few bucks by going with a lower bin, or maybe it's just underclocked to stretch out the battery life. 200MHz across 4 cores definitely makes a difference in benchmarks, especially with the amount of pixels they have to push on the 8.3 (1920x1200 makes it 1200p, not the standard 1080p of FHD). The Galaxy S3 might get 15000 in Antutu, but it also has less than half the pixels to push. My Galaxy Light, which is NOT a high performance device, also gets 15000-16000, but it has less than half the pixels of the S3. Remember, benchmarks aren't always an indicator of real-world performance.
EDIT: I haven't used the Verizon 8.3, but being a Verizon device, I imagine it's chocked full of the typical Verizon bloat. Check your processes to see if there's something unnecessary running in the background and delete that crap, or at least disable or Greenify it.
Click to expand...
Click to collapse
Good stuff here... Thanks!
Yeah, honestly I typically don't depend on benchmarks, but like to refer to them as a ballpark figure of sorts. For instance, on PC, I typically wouldn't be able to tell a single difference in terms of actual, real worl performance when comparing say 3 Ghz quad vs a 3.2 Ghz quad. But synthetic benchmarks might show one edging the other out by a significant enough margin to make it worth mentioning.
I ran Antutu on my Moto G - a 1.2 Ghz quad core Snapdragon 400 - and pulled like ~16600, but only got ¬15700 on the 1.5 Ghz quad core Snapdragon 600 found in the G Pad. Could be a software bug perhaps too. I also flashed the June 2014 Adreno drivers, but it hasn't seemed to make a difference in how this thing is scoring.
Edit: Okay, weird. Both Antutu & CPU-Z are reporting that I have a 8064 Snapdragon S4 Pro... NOT the 8064T Snapdragon 600. Would explain a lot. Spec sheets show the G Pad (presumably the V500/V510) having the 600. I used a hardware identifier app, but it reports I have the Snapdragon 600 on 8064 board. Now I'm wonderin if the VK810 ships with an S4 Pro instead of the 600 like the V500V510.
Kaiyo Droid said:
Take this for instance: I ran Antutu on my Moto G - a 1.2 Ghz quad core Snapdragon 400 - and pulled like ~16600, but only got ¬15700 on the 1.5 Ghz quad core Snapdragon S4 Pro found in the VK810. Could be a software bug perhaps too, but not sure what's going on. I also flashed the June 2014 GPU drivers, but it hasn't seemed to make a difference in how this is scoring.
Click to expand...
Click to collapse
Remember that the Moto G has less than half the pixels to drive than the VK810, so it takes less processing power to keep the framerates up and thus score well on the benchmarks. 3D games run far smoother (and with higher detail settings) on my Nexus 4 than on my 8.3 V500, most likely because of this same thing.
Still, your scores do seem low, and I would have expected the optimized binaries to improve things. If you look at this thread, others with the VK810 have been getting 19-20000 scores, although they also report some choppiness. I guess itt's possible you got a bum unit.
Planterz said:
Remember that the Moto G has less than half the pixels to drive than the VK810, so it takes less processing power to keep the framerates up and thus score well on the benchmarks. 3D games run far smoother (and with higher detail settings) on my Nexus 4 than on my 8.3 V500, most likely because of this same thing.
Still, your scores do seem low, and I would have expected the optimized binaries to improve things. If you look at this thread, others with the VK810 have been getting 19-20000 scores, although they also report some choppiness. I guess itt's possible you got a bum unit.
Click to expand...
Click to collapse
Thanks! I think I'm beginning to understand the difference in needing to push more pixels now.
Came across this thread too regarding someone's trouble identifying the CPU on their LG device using CPU-Z and Antutu: http://forum.xda-developers.com/showthread.php?t=2341947&page=2
Kaiyo Droid said:
Came across this thread too regarding someone's trouble identifying the CPU on their LG device using CPU-Z and Antutu: http://forum.xda-developers.com/showthread.php?t=2341947&page=2
Click to expand...
Click to collapse
What does CPU-Z say your VK810 has? I've looked around and can't find specifics on this anywhere that I can be sure of. Both LG and Verizon merely list it as a "1.5 GHz Quad-Core Processor, Qualcomm® Snapdragon™ Chipset". There's other sites that state that it's a Snapdragon 600, but this could just be an assumption they made because the V500/510 have the S600.
According to wikipedia, The S600 is sometimes marketed as an S4 pro. The Nexus 7 (2013) for example, is advertised as having an S4 Pro, but apparently it's actually in the S600 series. On my V500, CPU-Z gives me a model number of APQ8064T. If CPU-Z gives you a model number, you might be able to track down exactly what it has on the wikipedia page I linked to.
Planterz said:
What does CPU-Z say your VK810 has? I've looked around and can't find specifics on this anywhere that I can be sure of. Both LG and Verizon merely list it as a "1.5 GHz Quad-Core Processor, Qualcomm® Snapdragon™ Chipset". There's other sites that state that it's a Snapdragon 600, but this could just be an assumption they made because the V500/510 have the S600.
According to wikipedia, The S600 is sometimes marketed as an S4 pro. The Nexus 7 (2013) for example, is advertised as having an S4 Pro, but apparently it's actually in the S600 series. On my V500, CPU-Z gives me a model number of APQ8064T. If CPU-Z gives you a model number, you might be able to track down exactly what it has on the wikipedia page I linked to.
Click to expand...
Click to collapse
As you stated, many of the places that list specs for the LTE model of the G Pad 8.3 simply mirror the ones found for the V500/V510; incorrectly specifying the SoC as the S600 clocked at 1.7Ghz (which we know isn't true). Probably simply assumed they would be the same, only one would feature LTE connectivity.
http://pdadb.net/index.php?m=specs&id=5693&c=lg_vk810_g_pad_8.3_4g_lte
http://www.phonearena.com/phones/LG-G-Pad-8.3_id8141 (scroll to the bottom under Variants)
CPU-Z reports that my VK810 has an S4 Pro, APQ8064. However, the hardware identifier app states it's an S600 clocked @ 1.5 Ghz, APQ8064"T". The Antutu Benchmark and Tester apps also say it's the APQ9064T clocked @ 1.5Ghz. Maybe it's a software side miscalculation with CpU-Z, yet it still doesn't explain why CPU-Z can properly identify the S600 in V500/V510's, but not the VK810.
Qualcomm's own website for the Snapdragon 600 specifically cites the LG G Pad 8.3 LTE (VK810) as having the S600 APQ8064T under the hood (as well as the Galaxy S4 and HTC One). Oddly, the V500/V510 aren't listed on that page, only the LTE variant. Link: http://www.qualcomm.com/snapdragon/processors/600
Other sources:
http://www.devicespecifications.com/en/model/122a2b68
http://www.gsmarena.com/lg_g_pad_8_3_lte-6173.php
I used SetCPU to change the govenor from On Demand to Performance, ran Antutu twice and scored ~18800 both times. I haven't tried disabling thermal throttling, because frankly, I haven't been able to find out how.
Now to figure out why my unit is benching well below other users' VK810's.
Fixed! (Well, somewhat...)
{
"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"
}
So, here's what I ended up doing:
I am on 4.4.2, rooted.
I unistalled SetCPU, and installed the "Trickster MOD Kernel Settings" app (Google Play: https://play.google.com/store/apps/details?id=com.bigeyes0x0.trickstermod&hl=en )
In Trickster MOD, I did the following:
Under the General menu, I set the governor to "Interactive".
Under the Specific Menu, I changed the following:
Temperature Throttle = On (default is OFF)
GPU Governor = On Demand or Performance (either one seemed to work, default is On Demand)
GPU Max Frequency = 400 (default on mine was 320 for some reason)
Once you set everything up, reboot the tablet.
Recall that in the beginning, with no tweaks, I was scoring around 16000 in Antutu Benchmark.
With these minor tweaks, I've benchmarked it several times and now consistently score between 21000 to 22000 in Antutu.
The pixel XI and the pixel are packed with snadragon 821 chipset wich supposed to be clocked at 2*2.35 kryo & 2*2.0 kryo but both pixel phones are clocked at 2*2.15 kryo & 2*1.6 Kryo which is exactly the same as SD820 on Lg G5 and the s7 so if someone knows what is the difference between the cpu in the pixel phones and the regular snapdragon 820 please write it down
From what I have read the 821 is a 820. The 821 is just higher binned 820. When they make chips they are not all the same. Some just are a little more efficient than others do to very minor differences in the chips. So a high binned 820 can handle a higher clock speed while using less power are turned into 821.
So Google decided they wanted to go with the 821 because it is more power efficient than a 820. But it seems Google thinks the speed of the 820 is fast enough to provide a good snappy user experience. So they are doubling down on efficiency by clocking these high binned chips down to the same as the 820. So say the 821 is 5% more efficient at stock speed over the 820. The 821 might be 10% more efficient at the same clock speed at the 820 while delivering the same speed as the 820. So they are sort of doubling down on efficiency over performance.
From the hands on I have seen everyone has described the phone as very fast. This is likely due to Google optimizing Android to run on the pixels hardware. Much like Apple does with the iPhone. Also the Pixel has some hardware features that might not show up on a regular spec sheet. It has some improved touch screen latency and faster storage. Because of these factors Google decided they don't need the extra performance of the 821 but instead want to utilize it's efficiency.
TLDR Google is going all in on the Pixel proving a very fast user experience while being power efficient!
So in theory once kernel source has been released we can just OC it back to "stock" frequency and get even faster performance with a hit to battery life.
I have the OP3 and the phone is clocking to max. frequency very rarely anyway. So there is no reason to clock it down for better efficiency.
So basically the pixel xl nd the pixel have snadragon 820 with a different name and better efficiency, as a result the gaming performance is the same as on the lg g5 or the s7 for example, these pixel devices arent worth the extra 200$
ramqashou said:
So basically the pixel xl nd the pixel have snadragon 820 with a different name and better efficiency, as a result the gaming performance is the same as on the lg g5 or the s7 for example, these pixel devices arent worth the extra 200$
Click to expand...
Click to collapse
The smaller Pixel has the potential to out do both of those phones and the Pixel XL in gaming since it has a native resolution of 1080p. The lower the resolution, the higher frames per second possible in games when using the same SoC, assuming the game is made to run at your phones native resolution.
ramqashou said:
So basically the pixel xl nd the pixel have snadragon 820 with a different name and better efficiency, as a result the gaming performance is the same as on the lg g5 or the s7 for example, these pixel devices arent worth the extra 200$
Click to expand...
Click to collapse
In the current climate and with the 810 fiasco overshadowing can you really blame them for dialing it down? Perhaps the GPU is still clocked higher in the 821 and I'll take the efficiency as a perk. It's up to you what's worth $200 more but there are a few more bits less talked about included in the price.
---------- Post added at 07:11 AM ---------- Previous post was at 07:09 AM ----------
mixedguy said:
The smaller Pixel has the potential to out do both of those phones and the Pixel XL in gaming since it has a native resolution of 1080p. The lower the resolution, the higher frames per second possible in games when using the same SoC, assuming the game is made to run at your phones native resolution.
Click to expand...
Click to collapse
I'd rather have 1080p at 60FPS than 2k at 30FPS on a screen that size, however I think most games, at least the big titles, have adjustable resolution so I think the only difference will be battery draw.
Hoodeddeathman said:
I'd rather have 1080p at 60FPS than 2k at 30FPS on a screen that size, however I think most games, at least the big titles, have adjustable resolution so I think the only difference will be battery draw.
Click to expand...
Click to collapse
I agree, I wasn't aware you could choose your resolution in mobile phone games as I don't really play demanding games on my phone, I assumed it was like mainstream game consoles where the developer predetermines the resolution or just sets it to use the native res by default.
I play games on PC, so it's pretty cool that you can change the resolution on mobile phone games like you can on PC games.
mixedguy said:
I agree, I wasn't aware you could choose your resolution in mobile phone games as I don't really play demanding games on my phone, I assumed it was like mainstream game consoles where the developer predetermines the resolution or just sets it to use the native res by default.
I play games on PC, so it's pretty cool that you can change the resolution on mobile phone games like you can on PC games.
Click to expand...
Click to collapse
As I understand it android has the capability and it's up to the devs to implement. The game can be rendered at whatever resolution and will then be upscaled. for example Warhammer Freeblade allows you to select which resolution to use and texture qualities just as you would in most PC games at the risk of losing frames however Need For Speed No Limits selects a pre-defined profile depending on device.
As I said, underclocking doesn't automatically mean better effiency... If you would have a 820 phone you would know that. I experimented a lot with different CPU settings on my One Plus 3 and underclocking is not worth it because it only cuts of performance but does NOT increase effiency because your CPU is using max frequency like 1% of the runtime anyway... In more than two days 2,15 GHz on the big cluster was used only 49s on my OP3.
And that the 821 reaches a higher frequency doesn't automatically mean that the CPU has a higher quality. I know it would be possible that the 820s are only bad 821s that don't surpass quality tests but I don't think so because the 820 was released much earlier. Usually it goes the other way around, like on GPUs. Nvidia first releases the very high end models and then sells the crappy GPUs in the lower end models. I don't think that Qualcomm is like, hey we are picking out all really good 820s and pile them up to sell them as 821s... A 821 could be better and more efficient but it's not necessarily true. A good 820 could still be as good or even better than a 821, regarding effiency. Also think about AMD Processors a few years ago, whole cores where unlock able and there was still room for OC if you were lucky.
Gerrit507 said:
As I said, underclocking doesn't automatically mean better effiency... If you would have a 820 phone you would know that. I experimented a lot with different CPU settings on my One Plus 3 and underclocking is not worth it because it only cuts of performance but does NOT increase effiency because your CPU is using max frequency like 1% of the runtime anyway... In more than two days 2,15 GHz on the big cluster was used only 49s on my OP3.
Click to expand...
Click to collapse
When talking about efficiency I'm referring more to undervolting as appose to underclocking, it may be the case that they have chosen those frequencies because the 821 steps up in voltage beyond that point thus increasing power consumption and heat. We'll have to wait and see how the Pixel performs, but if that underclock means the thermal load is capped lower we will also see less throttling, ideal for daydream.
As an example I would refer to overclocking desktop CPUs, the architecture is different but how it responds to heat and power is not. beyond a certain frequency the CPU requires exponentially more power and generates exponentially more heat the higher you go.
http://m.gsmarena.com/google_pixel_xl_benchmark_doesnt_show_performance_improvement-news-20927.php
This benchmark proves to all those who insist that the chipset in the pixel phones is better than the original snadragon 820
It might be only to reduce the heat. The battery efficiency is, IMO, very marginal.
But I will surely put back the 2.4GHz on mine.
firewave said:
It might be only to reduce the heat. The battery efficiency is, IMO, very marginal.
But I will surely put back the 2.4GHz on mine.
Click to expand...
Click to collapse
For some reason idont believe in overclocking, cuz it's beyond the device capabilities and it might cause some problems.
This seems a very big piece of marketing by Google. It isn't really an SD 821, its an SD820.
The 821 only has a 10% performance increase when clocked at its max frequency, so even if Google did leave it at its max frequency, a 10% increase would be barely noticeable, if noticeable at all in real world use.
The 821 does have some features that aren't available on the 820, which is why Google probably chose the 821 over the 820. I found this info about two important features for the 821, that's not found in the 820 and quoted it below.
"One of the main reason why Google used the Snapdragon 821 in the Pixel phones is the Snapdragon VR SDK (Software Development Kit). This is entirely unavailable with the Snapdragon 820. The new SDK comes with advanced VR toolset to give the developers broad access to the internal architecture of the Snapdragon 821 chipset. This is extremely useful and fully compatible with Google Daydream platform. The VR SDK helps in the rendering of cutting-edge visual and audio which helps in state of the art Virtual Reality experience."
"Another important thing which is unknown for most people is about the camera improvements brought by the MSM8996 Pro. The SoC can simultaneously use two phase detectors for significant improvement in focussing quality and time. On the contrary, the Snapdragon 820 or MSM8996 only supports single PDAF (Phase Detecting Auto Focus) systems. The newer chipset extends the range of laser autofocus technology. This will substantially boost the laser-assisted autofocus systems of upcoming smartphones."
ramqashou said:
For some reason idont believe in overclocking, cuz it's beyond the device capabilities and it might cause some problems.
Click to expand...
Click to collapse
That's a very incorrect statement. The kernel determines the clock speed. Google could choose something like 0.5GHz if they were so inclined. The phone would run like ****, but in your eyes, the device is not "capable" of anything faster. It sounds like Google purposely underclocked these. If nothing else, you are absolutely 100% fine to clock it back to the speed that Qualcomm, the OEM of the chipset, intended it to run at. True overclocking can present problems, but I have overclocked my CPUs, RAM, and GPUs for YEARS with no issues and reaped plenty of extra benefits in terms of performance. I used to do it on my smartphones too, but it is pointless and wastes battery for almost every use scenario.
Google specifically chose 2.15GHz instead of 2.4GHz as specified by Qualcomm, either due to heat issues or battery life benefit. I am going to guess they realized that their incredibly light and optimized software does not need a 2.4GHz CPU speed - hell, my 6P is faster with a SD 810 than my Note7 with an 820 in day to day use for a reason, that reason being stock Android is incredibly quick and efficient.
That is true from the chip standpoint. What you don't know, though, is if google/htc designed the heat removal system to handle the additional heat produced at full clock speeds without throttling...
Sent from my Nexus 6P using Tapatalk
JasonJoel said:
That is true from the chip standpoint. What you don't know, though, is if google/htc designed the heat removal system to handle the additional heat produced at full click speeds without throttling...
Sent from my Nexus 6P using Tapatalk
Click to expand...
Click to collapse
The phone being a uni-body aluminium shell should help with that. My 5X gets mega hot when I run games or for extended screen on times, but the back is plastic. I think using the whole surface of the phone as an additional heat-sink so to speak could help with heat dissipation.
Either way - I hope someone tries to OC it back to "stock" qualcomm speeds. I will certainly try to see the results, that is, if custom kernels can be a thing with the Pixel.
Nitemare3219 said:
That's a very incorrect statement. The kernel determines the clock speed. Google could choose something like 0.5GHz if they were so inclined. The phone would run like ****, but in your eyes, the device is not "capable" of anything faster. It sounds like Google purposely underclocked these. If nothing else, you are absolutely 100% fine to clock it back to the speed that Qualcomm, the OEM of the chipset, intended it to run at. True overclocking can present problems, but I have overclocked my CPUs, RAM, and GPUs for YEARS with no issues and reaped plenty of extra benefits in terms of performance. I used to do it on my smartphones too, but it is pointless and wastes battery for almost every use scenario.
Google specifically chose 2.15GHz instead of 2.4GHz as specified by Qualcomm, either due to heat issues or battery life benefit. I am going to guess they realized that their incredibly light and optimized software does not need a 2.4GHz CPU speed - hell, my 6P is faster with a SD 810 than my Note7 with an 820 in day to day use for a reason, that reason being stock Android is incredibly quick and efficient.
Click to expand...
Click to collapse
That's true i can't
Deny the power of stock android, but there are many other OEM custom skins that are well optimized and are plenty fast such as sense and Lg ux 5.0 and even the oxygen OS
Regarding the OnePlus throttling gate, you can read here:
Examining OnePlus' Performance Behaviour: Optimization or Misrepresentation?
www.anandtech.com
The OnePlus 9 seems to be throttling its performance in many popular apps
The OnePlus 9 Pro seems to be limiting the performance in popular apps like Chrome, Twitter, Zoom, WhatsApp, Facebook, Instagram, and more.
www.xda-developers.com
This is just stupid, I have been running a 7 pro for 4 years, it has been, hands down, one of the best smartphones I ever owned, performance has been great from day 1 to last day and it is still living with a relative of mine, and he is supper happy with it, I have play a lot, like A LOT and it has worked flawlessly every single time.
I got a 9 pro in october past year and it has been a pain in the ass, the poor performance has been annoyingly usual, the only way to play 120FPS in some games is a magisk module intended for a Samsung F62 which I can't figure out why (because I have no knowledge about this) but it makes some game run 120FPS, however performance is really bad, there is frame drops in some points where my couple's nord 2 demolishes with no issue... I can't figure out why a nord 2 outperforms my 9 pro (i know nord 2 is 90Hz but it is much more solid 90FPS being lower tier).
Let me explain, my couple nord 2 remains 90FPS where my 9 pro gets as low as 30FPS, it is noticeable and annoying.
Is there some sort of magisk module or something "comfortable" for newbies which can be enabled or disabled at taste? I don't mind rebooting phone (I do every time i want to play some games with module I mentioned xd)
I did not tinker with kernels (not yet, did in the past with my former LG G2 and Galaxy S3 but it was quite different). I would like to remain in the Oxygen OS path for now too.
I tried lots of things, AI booster for magisk, also Adreno drivers and nothing seems to work, just performance scaling which seems to lower freq in order to improve battery life, but never higher performance (high performance mode in battery settings seems to do nohing, also tried)
Looking for this as well, as well as a module to disable annoying Doze by the SystemUI. I tried a few modules but the devices still dozes.
SnoopyFTW said:
Regarding the OnePlus throttling gate, you can read here:
Examining OnePlus' Performance Behaviour: Optimization or Misrepresentation?
www.anandtech.com
The OnePlus 9 seems to be throttling its performance in many popular apps
The OnePlus 9 Pro seems to be limiting the performance in popular apps like Chrome, Twitter, Zoom, WhatsApp, Facebook, Instagram, and more.
www.xda-developers.com
This is just stupid, I have been running a 7 pro for 4 years, it has been, hands down, one of the best smartphones I ever owned, performance has been great from day 1 to last day and it is still living with a relative of mine, and he is supper happy with it, I have play a lot, like A LOT and it has worked flawlessly every single time.
I got a 9 pro in october past year and it has been a pain in the ass, the poor performance has been annoyingly usual, the only way to play 120FPS in some games is a magisk module intended for a Samsung F62 which I can't figure out why (because I have no knowledge about this) but it makes some game run 120FPS, however performance is really bad, there is frame drops in some points where my couple's nord 2 demolishes with no issue... I can't figure out why a nord 2 outperforms my 9 pro (i know nord 2 is 90Hz but it is much more solid 90FPS being lower tier).
Let me explain, my couple nord 2 remains 90FPS where my 9 pro gets as low as 30FPS, it is noticeable and annoying.
Is there some sort of magisk module or something "comfortable" for newbies which can be enabled or disabled at taste? I don't mind rebooting phone (I do every time i want to play some games with module I mentioned xd)
I did not tinker with kernels (not yet, did in the past with my former LG G2 and Galaxy S3 but it was quite different). I would like to remain in the Oxygen OS path for now too.
I tried lots of things, AI booster for magisk, also Adreno drivers and nothing seems to work, just performance scaling which seems to lower freq in order to improve battery life, but never higher performance (high performance mode in battery settings seems to do nohing, also tried)
Click to expand...
Click to collapse
The links you posted, those articles are not RELEVANT to YOUR issue. The throttling you're experiencing is a thermal throttling behavior that is literally NORMAL and part of EVERY OEM/Phone.
If you're on a STOCK ROM, the only way to somewhat "kill" the default throttling behavior is to disable/freeze the BATTERY package (app) either via Titanium Backup or some other app. That will prevent the phone from throttling based on the SKIN/Battery temperature.
Otherwise, Custom ROMs based on either OOS12 or OOS13 don't have any of that stock OEM throttling behavior.
JohnTheFarm3r said:
The links you posted, those articles are not RELEVANT to YOUR issue. The throttling you're experiencing is a thermal throttling behavior that is literally NORMAL and part of EVERY OEM/Phone.
If you're on a STOCK ROM, the only way to somewhat "kill" the default throttling behavior is to disable/freeze the BATTERY package (app) either via Titanium Backup or some other app. That will prevent the phone from throttling based on the SKIN/Battery temperature.
Otherwise, Custom ROMs based on either OOS12 or OOS13 don't have any of that stock OEM throttling behavior.
Click to expand...
Click to collapse
Is it really thermal throttling? I mean, phone does not get any warmer than usual and for example, Brawl Stars is capped at 60 FPS like most games in BBK phones (but Brawl stars worked 120FPS flawless some time ago, not much, a year ago or so I had a Realme GT Neo 2 which ran games buttersmooth and has basically the same software, color OS, which now has the same issue, games capped to 60 FPS and crap performance)
That is not normal behaviour because Brawl Stars devs claimed the game is fully unlocked and if it works below screen refresh is because of the phone software, that's the "throttling" I mean, don't know if it is because of the famous whitelisting of Oppo.
What I mean is in previous software games worked properly and now performance is a joke (I mention realme GT Neo 2 because is what I could test in the past, I loved it, that's why is a bummer getting a "better phone" which works much worse).
Get an Xbox and solve all your game problems!
TheGhost1951 said:
Get an Xbox and solve all your game problems!
Click to expand...
Click to collapse
An Xbox? Are you from USA?
I just want my phone works as it should, I don't care about battery.
Also tried disabling battery app and no difference, same behaviour, I'm in OOS13 latest update.
Maybe I should give a try to custom ROM...
SnoopyFTW said:
Is it really thermal throttling? I mean, phone does not get any warmer than usual and for example, Brawl Stars is capped at 60 FPS like most games in BBK phones (but Brawl stars worked 120FPS flawless some time ago, not much, a year ago or so I had a Realme GT Neo 2 which ran games buttersmooth and has basically the same software, color OS, which now has the same issue, games capped to 60 FPS and crap performance)
That is not normal behaviour because Brawl Stars devs claimed the game is fully unlocked and if it works below screen refresh is because of the phone software, that's the "throttling" I mean, don't know if it is because of the famous whitelisting of Oppo.
What I mean is in previous software games worked properly and now performance is a joke (I mention realme GT Neo 2 because is what I could test in the past, I loved it, that's why is a bummer getting a "better phone" which works much worse).
Click to expand...
Click to collapse
it is because of thermal throttling, literally. The phone doesn't get warmer because it's being limited, hence the point of thermal throttling.
The refresh rate - FPS - in a game depends solely on the devs. If the game has in one device for example support for 120 FPS, and it doesn't have in another device, it's not because of the device but because the developers either didn't "whitelist" that specific device or weren't updated in a longer time.
JohnTheFarm3r said:
it is because of thermal throttling, literally. The phone doesn't get warmer because it's being limited, hence the point of thermal throttling.
The refresh rate - FPS - in a game depends solely on the devs. If the game has in one device for example support for 120 FPS, and it doesn't have in another device, it's not because of the device but because the developers either didn't "whitelist" that specific device or weren't updated in a longer time.
Click to expand...
Click to collapse
Hmm...
With the magisk module I mentioned device name changes from LE2123 to SO-52A or something similar, makes also work codm 120 FPS pretty good, much better than brawl stars which is much weaker in terms of graphics quality and detail.
Is there any way to spoof model name to test this? If so devs are lying big time. When launched 9 pro they announced collab with many games, even hypertouch feature which is gone and for what seems now touch sampling rate is just 120Hz from the 360 advertised (monitoring with apps say so).
Also tried to disable battery app and no difference, behaviour is the same.
SnoopyFTW said:
Hmm...
With the magisk module I mentioned device name changes from LE2123 to SO-52A or something similar, makes also work codm 120 FPS pretty good, much better than brawl stars which is much weaker in terms of graphics quality and detail.
Is there any way to spoof model name to test this? If so devs are lying big time. When launched 9 pro they announced collab with many games, even hypertouch feature which is gone and for what seems now touch sampling rate is just 120Hz from the 360 advertised (monitoring with apps say so).
Also tried to disable battery app and no difference, behaviour is the same.
Click to expand...
Click to collapse
Disabling the battery app is for the throttling workaround, not to increase the frames in the game that don't support it on your device. lol
Anyways, blame the devs...They're the ones that need to add support for the specific device to go above the XY FPS.
The smartphone wasn't intended to be a gaming device. Chipset and battery not rated for it. Get an Xbox or some similar device for gaming that it was intended for! Just my opinion.
You can test the throttling using a CPU/GPU Throttle Testing app. The OP 9 starts throttling at a low temperature, I can notice FPS dropping in games starting at 38 degrees.
I use a cooling fan attached to my OP to keep the phone at a constant 32 degrees. I also bought a heat dispersion case to help further.
Also using Konabess to undevolt the GPU also helps a bit.
TheGhost1951 said:
The smartphone wasn't intended to be a gaming device. Chipset and battery not rated for it. Get an Xbox or some similar device for gaming that it was intended for! Just my opinion.
Click to expand...
Click to collapse
I'm not asking for it to perform in graphic demanding games, I'm telling you even a redmi note 10 pro (that thing is actually not rated for anything) outperforms the 9 pro in light gaming because it is not capped.
Blame the devs, blame OnePlus, I just want to know if there is any solution, I purchased the phone a few months ago and I'm finding stupid behaviours an "allegedly" flagship should not have.
Jeez even my couple Nord 2 is much better in games, that hurts...
And yeah. Get an Xbox blablabla, I do own a PS4, a powerful laptop and a powerful desktop, I just want my stupid phone to do what it should do or at least know why or if it can be solved, this is not r/xbox.
This is an issue many people complain about but OnePlus ever stated anything.
For what I can see my options are getting a cheap 120HZ phone or custom ROM (I saw a blackshark 4 for 200EUR which runs games fine)
JohnTheFarm3r said:
Disabling the battery app is for the throttling workaround, not to increase the frames in the game that don't support it on your device. lol
Anyways, blame the devs...They're the ones that need to add support for the specific device to go above the XY FPS.
Click to expand...
Click to collapse
The thing is when released the 9 pro, OnePlus brand had collabs with many games, even hypertouch, all of it disappeared into thin air...
That was the main reason for me to purchase, I had a great experience with my former 7 pro and I couldn't even think this could be a thing.
https://twitter.com/i/web/status/1367122914961670156
This thread is of the main dev of the game I'm talking about. He claims the brand has to whitelist app in order to make it work 120HZ/120FPS.
Is there no way to modify that "whitelist"? I've read sometimes about this but rarely found anything (anything users can do about it lmao) and we all know how lazy oppo is with this things, there can be a thousand posts in OnePlus community ther will never answer.
SnoopyFTW said:
The thing is when released the 9 pro, OnePlus brand had collabs with many games, even hypertouch, all of it disappeared into thin air...
That was the main reason for me to purchase, I had a great experience with my former 7 pro and I couldn't even think this could be a thing.
https://twitter.com/i/web/status/1367122914961670156
This thread is of the main dev of the game I'm talking about. He claims the brand has to whitelist app in order to make it work 120HZ/120FPS.
Is there no way to modify that "whitelist"? I've read sometimes about this but rarely found anything (anything users can do about it lmao) and we all know how lazy oppo is with this things, there can be a thousand posts in OnePlus community ther will never answer.
Click to expand...
Click to collapse
This is a GAME SIDE issue, not a phone-side issue. There is no whitelist on a phone that will enable you the unsupported FPS.
SnoopyFTW said:
An Xbox? Are you from USA?
I just want my phone works as it should, I don't care about battery.
Also tried disabling battery app and no difference, same behaviour, I'm in OOS13 latest update.
Maybe I should give a try to custom ROM...
Click to expand...
Click to collapse
If you want to try a custom rom, Extended is great as daily driver and for gaming as well
Well, if XBox isn't available then look for a Microsoft Play Station instead!
JohnTheFarm3r said:
This is a GAME SIDE issue, not a phone-side issue. There is no whitelist on a phone that will enable you the unsupported FPS.
Click to expand...
Click to collapse
You are totally correct my friend... I have been tinkering with vmos app, I created a virtual machine in my own phone (lmao sounds like science fiction, but damnit this 888 has muscle) and it is just because devs do it intentionally.
I tried different phone makes and models... Casually Snapdragon flagship cap the game to 60FPS, but NORD devices make it run 120 FPS (you can set FPS settings in the VM option).
This is hilarious. Seems to me OnePlus just broke some contract conditions and this is devs punishment... Makes no sense, I have to test further but this is what I found for now. My doubt is if I might get a ban for using the game in a virtual machine... This is not relevant for this post!
Hope it helps someone, for exclusive games working just 120HZ in some devices you can spoof model name without root
SnoopyFTW said:
You are totally correct my friend... I have been tinkering with vmos app, I created a virtual machine in my own phone (lmao sounds like science fiction, but damnit this 888 has muscle) and it is just because devs do it intentionally.
I tried different phone makes and models... Casually Snapdragon flagship cap the game to 60FPS, but NORD devices make it run 120 FPS (you can set FPS settings in the VM option).
This is hilarious. Seems to me OnePlus just broke some contract conditions and this is devs punishment... Makes no sense, I have to test further but this is what I found for now. My doubt is if I might get a ban for using the game in a virtual machine... This is not relevant for this post!
Hope it helps someone, for exclusive games working just 120HZ in some devices you can spoof model name without root
Click to expand...
Click to collapse
Try your own phone model in the virtual machine. What happens?
Arealhooman said:
Try your own phone model in the virtual machine. What happens?
Click to expand...
Click to collapse
If I set up OnePlus LE2123/20/25 game runs 60FPS.
If I set for instance, my couple nord 2 DN2103 runs 120FPS same settings, just device name and brand changed!
Presets for S20, Xiaomi 10, vivo x50 all run 120 FPS.
Which for me is big yikes It is clearly devs fault, there has to be something odd...
SnoopyFTW said:
You are totally correct my friend... I have been tinkering with vmos app, I created a virtual machine in my own phone (lmao sounds like science fiction, but damnit this 888 has muscle) and it is just because devs do it intentionally.
I tried different phone makes and models... Casually Snapdragon flagship cap the game to 60FPS, but NORD devices make it run 120 FPS (you can set FPS settings in the VM option).
This is hilarious. Seems to me OnePlus just broke some contract conditions and this is devs punishment... Makes no sense, I have to test further but this is what I found for now. My doubt is if I might get a ban for using the game in a virtual machine... This is not relevant for this post!
Hope it helps someone, for exclusive games working just 120HZ in some devices you can spoof model name without root
Click to expand...
Click to collapse
So what are you saying exactly: if we spoof our device to a different one, we get more fps?