Related
I have tried 3 different S7 and one of my friends S7edge and all 4 have been inferior in their quality and visuals in gear vr (this also seems at least to me to carry over to non vr use as well) So i decided to capture video and show a comparison between the s7 and note 5 in gear vr especially while playing gunjack, but this happens in all apps and games to some degree but its noticeably worse in gunjack. I didnt play as long with the s7 because the flicker makes me nauseous.
S7
Note5
Also I did two more videos showing how the GS7 fails to render edges as smooth and clear as the note5, these are best viewed on mobile devices but are also easily noticed on pc platforms. Please forgive the watermark at the beginning. Please full screen to fairly compare the two.
S7
Note5
I have even compared the S6 and it performs better graphically although it overheats faster.
Any thoughts?
strokercrate said:
I have tried 3 different S7 and one of my friends S7edge and all 4 have been inferior in their quality and visuals in gear vr (this also seems at least to me to carry over to non vr use as well) So i decided to capture video and show a comparison between the s7 and note 5 in gear vr especially while playing gunjack, but this happens in all apps and games to some degree but its noticeably worse in gunjack. I didnt play as long with the s7 because the flicker makes me nauseous.
Also I did two more videos showing how the GS7 fails to render edges as smooth and clear as the note5, these are best viewed on mobile devices but are also easily noticed on pc platforms. Please forgive the watermark at the beginning. Please full screen to fairly compare the two.
I have even compared the S6 and it performs better graphically although it overheats faster.
Any thoughts?
Click to expand...
Click to collapse
Exynos or Snapdragon? Probably not as optimized for the S7 yet (most games came out long before the S7 was released).
Snapdragon. This happens even in oculus home, also the stair stepping wavy boarders are out of game. I video chatted samsung last night and showed them the videos, they are pushing it to some tech that looks into and breaks devices just to fix problems, and he he couldn't figure it out he woukd push it further up.
I have heard exynos is running pretty smooth for the most part... I have done some benchmarks with different tools that seem to indicate that while the cpu and gpu of the snapdragon is much faster than the note5 and s6 exynos, the I/O gets rated much slower. Could it be bottlenecked?
strokercrate said:
I have tried 3 different S7 and one of my friends S7edge and all 4 have been inferior in their quality and visuals in gear vr
Any thoughts?
Click to expand...
Click to collapse
I have a S7 edge and I play Gunjack. My S7E in no way has that jerky movement displayed on the video. If it did I wouldn't use it. Mine work about the same as your Note 5 example. I'm curious if this is happening with any other S7 owners. I had also had a S6. I couldn't use my Gear for more then 10 minutes with the S6 before I got an over heating warning. The Gear VR resolution is sharper on my S7E then it was on my S6.
Sent from my SM-G935V using XDA-Developers mobile app
Eddie Hicks said:
I have a S7 edge and I play Gunjack. My S7E in no way has that jerky movement displayed on the video. If it did I wouldn't use it. Mine work about the same as your Note 5 example. I'm curious if this is happening with any other S7 owners. I had also had a S6. I couldn't use my Gear for more then 10 minutes with the S6 before I got an over heating warning. The Gear VR resolution is sharper on my S7E then it was on my S6.
Sent from my SM-G935V using XDA-Developers mobile app
Click to expand...
Click to collapse
There are a those in both the gs7 and gs7edge boards on xda mentioning this, and some on the galaxy gear board. Over at the oculus forums some are also mentioning they have similar issues plus a friend of mine is having it as well. I know at&t and sprint variants have this. Maybe Verizon doesnt?
strokercrate said:
There are a those in both the gs7 and gs7edge boards on xda mentioning this, and some on the galaxy gear board. Over at the oculus forums some are also mentioning they have similar issues plus a friend of mine is having it as well. I know at&t and sprint variants have this. Maybe Verizon doesnt?
Click to expand...
Click to collapse
I'll search around to see who else is experiencing this issue. I don't have a Note 5 to compare, and viewing comparisons on Youtube video's is helpful but not the same as doing it yourself with the actual devices. Maybe in some ways the Note 5 works better with Gear VR, the only thing I can say for sure is I do not have anything close to that herky, jerky movement on my S7E that was displayed on that video. I learned over the years that even though you may have the exact same device there could be differences in how they work. For example, I have an issue of an overly sensitive touch screen on my S7E. It is so annoying I considered returning it. There are many others experiencing that issue. But most don't have it or don't consider it a problem. So maybe I'm the lucky one in regards to performance with my Gear VR.
http://forum.xda-developers.com/mobile-vr/gear-vr/choppy-video-s7-edge-t3359051
Eddie Hicks said:
I'll search around to see who else is experiencing this issue. I don't have a Note 5 to compare, and viewing comparisons on Youtube video's is helpful but not the same as doing it yourself with the actual devices. Maybe in some ways the Note 5 works better with Gear VR, the only thing I can say for sure is I do not have anything close to that herky, jerky movement on my S7E that was displayed on that video. I learned over the years that even though you may have the exact same device there could be differences in how they work. For example, I have an issue of an overly sensitive touch screen on my S7E. It is so annoying I considered returning it. There are many others experiencing that issue. But most don't have it or don't consider it a problem. So maybe I'm the lucky one in regards to performance with my Gear VR.
Click to expand...
Click to collapse
Just below this thread in gear vr board there is someone with an s7edge same issue.
I agree some are superior to others. But still shouldnt be this bad with so many. Including the issue you have with the sensitivity. Mine seems not sensitive enough i have to at times double touch things sometimes triple touch. My wife has bemoan that she hates her because its to sensitive opening stuff she has not touched and clicking while scrolling
I feel the s7/s7edge is the most buggy phone i have ever used from samsung and im a samsung fan boy and dislike the new flagships
http://forum.xda-developers.com/mobile-vr/gear-vr/choppy-video-s7-edge-t3359051
strokercrate said:
Mine seems not sensitive enough i have to at times double touch things sometimes triple touch. My wife has bemoan that she hates her because its to sensitive opening stuff she has not touched and clicking while scrolling
Click to expand...
Click to collapse
I'm in the same boat as your wife. I'm constantly opening apps and links inadvertently. I had a S6 for about 4 months. I got rid of it because of a very poor battery life. I didn't have that screen sensitivity issue with my S6. Maybe they rushed these S7's to market too soon. There are too many inconsistencies with performance.
Eddie Hicks said:
I'm in the same boat as your wife. I'm constantly opening apps and links inadvertently. I had a S6 for about 4 months. I got rid of it because of a very poor battery life. I didn't have that screen sensitivity issue with my S6. Maybe they rushed these S7's to market too soon. There are too many inconsistencies with performance.
Click to expand...
Click to collapse
Yes i agree, i had two and both have issues the first really bad so i returned it. Now my wifes the one so sensitive is developing issues where it black screens and wont respond nor power down unless i hard reset with volume down and power for 10 seconds. I guess another going back.
So i decided to actually monitor how much of the cpu had to be used to run both oculus and also gunjack and compare between note5 and s7, again the results show that the note5 has to use less processing power then does the s7 whats interesting is that I decided to do a 2 minute log of an idle device on both as well, and the s7 has to use more processing power for the same system apps than does the note 5.
S7
{
"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"
}
Note5
strokercrate said:
So i decided to actually monitor how much of the cpu had to be used to run both oculus and also gunjack and compare between note5 and s7, again the results show that the note5 has to use less processing power then does the s7 whats interesting is that I decided to do a 2 minute log of an idle device on both as well, and the s7 has to use more processing power for the same system apps than does the note 5.
Click to expand...
Click to collapse
Probably calculating the percentages differently as the S7 snapdragon only has four cores, so if it uses two cores it's 50% on the S7 and 25% on the Note 5. CPU really isn't what you should be looking at anyhow as the GPU is doing much of the work. There really doesn't have to be any other explanation at this point than that the apps for oculus aren't optimized for the Snapdragon SoCs just yet and it is probably going to take them awhile to sort it out. Would love to try it out on my Exynos version to see if it's a SoC-issue or something else, but the game is really expensive for something I'd just use to test it out.
Have you tried to contact the developers of Gunjack?
Toss3 said:
Probably calculating the percentages differently as the S7 snapdragon only has four cores, so if it uses two cores it's 50% on the S7 and 25% on the Note 5. CPU really isn't what you should be looking at anyhow as the GPU is doing much of the work. There really doesn't have to be any other explanation at this point than that the apps for oculus aren't optimized for the Snapdragon SoCs just yet and it is probably going to take them awhile to sort it out. Would love to try it out on my Exynos version to see if it's a SoC-issue or something else, but the game is really expensive for something I'd just use to test it out.
Have you tried to contact the developers of Gunjack?
Click to expand...
Click to collapse
Gunjack is the most pronounced. It happens everywhere in oculus, including home. Thats why one video shows the lack of smooth boarders on boxes just in oculus on the s7.
Plus i notice the s7 lags everywhere. Even doing common every day task, compared to previous s6 and note devices. Have ran test showing the snap dragon variants of s7 and edge has less i/o scores then previous gens as well.
strokercrate said:
Gunjack is the most pronounced. It happens everywhere in oculus, including home. Thats why one video shows the lack of smooth boarders on boxes just in oculus on the s7.
Plus i notice the s7 lags everywhere. Even doing common every day task, compared to previous s6 and note devices. Have ran test showing the snap dragon variants of s7 and edge has less i/o scores then previous gens as well.
Click to expand...
Click to collapse
Which tests? Care to post your results? Never seen anyone with lower scores than previous generation. Have you tried a full reset and updating it to the latest apd1 firmware? Sounds like something is wrong with your phone. Maybe a faulty nand-chip or something.
Toss3 said:
Which tests? Care to post your results? Never seen anyone with lower scores than previous generation. Have you tried a full reset and updating it to the latest apd1 firmware? Sounds like something is wrong with your phone. Maybe a faulty nand-chip or something.
Click to expand...
Click to collapse
Tested 3 galaxy s7 and one galaxy edge with three different benchmarks tools showing lower i/o on the s7 and s7edge than note 5 and s6, processor and gpu did score higher but lower i/o means bottleneck. I believe thats the reason for the jitter and shudder when browsing playing vr. Most of my post have been over on oculus forums with pictures. https://forums.oculus.com/community/discussion/33459/galaxy-s7-gear-vr-issues#latest
strokercrate said:
Tested 3 galaxy s7 and one galaxy edge with three different benchmarks tools showing lower i/o on the s7 and s7edge than note 5 and s6, processor and gpu did score higher but lower i/o means bottleneck. I believe thats the reason for the jitter and shudder when browsing playing vr. Most of my post have been over on oculus forums with pictures. https://forums.oculus.com/community/discussion/33459/galaxy-s7-gear-vr-issues#latest
Click to expand...
Click to collapse
Really wouldn't trust a beta benchmark (you only used Vellamo for some reason), and in 3DMark the S7 Edge gives you a higher score, but runs a bit hotter (still means it is faster). Why not try Geekbench (which most people use to measure single and multithreaded performance)? And Androbench for storage. I mean I am getting even lower IO scores than you and I have no stuttering (but topping the benchmarks by a fair margin in Androbench and Geekbench, so I think Vellamo shouldn't be trusted).
EDIT: Even my Note 4 it outscoring my S7 in Vellamo, so yeah that should already tell you a lot about how trustworthy the numbers are.
EDIT2: https://www.reddit.com/r/GearVR/comments/4ak3nv/galaxy_s7_stutters_like_crazy_compared_to_s6/d11jomb
"Carmack said it is in fact due to the software being unoptimized. Said s7 launch was rushed to the gates and they were not ready."
So there's no real performance bottleneck just unoptimized software (still wouldn't explain why there aren't more complaints and a lot of people saying they aren't seeing any issues).
Toss3 said:
Really wouldn't trust a beta benchmark (you only used Vellamo for some reason), and in 3DMark the S7 Edge gives you a higher score, but runs a bit hotter (still means it is faster). Why not try Geekbench (which most people use to measure single and multithreaded performance)? And Androbench for storage. I mean I am getting even lower IO scores than you and I have no stuttering (but topping the benchmarks by a fair margin in Androbench and Geekbench, so I think Vellamo shouldn't be trusted).
EDIT: Even my Note 4 it outscoring my S7 in Vellamo, so yeah that should already tell you a lot about how trustworthy the numbers are.
EDIT2: https://www.reddit.com/r/GearVR/comments/4ak3nv/galaxy_s7_stutters_like_crazy_compared_to_s6/d11jomb
"Carmack said it is in fact due to the software being unoptimized. Said s7 launch was rushed to the gates and they were not ready."
So there's no real performance bottleneck just unoptimized software (still wouldn't explain why there aren't more complaints and a lot of people saying they aren't seeing any issues).
Click to expand...
Click to collapse
I did run geekbench but it doesn't show i/o, everyone that shows i/o shows a considerable less on s7, quadrant is another that does and antutu, and those two show ram as better in the note 5.
Never ran storage on s7, so will now.
What variant do you have?
The s7 seems to have the largest variation in performance between multiple devices i have seen based on response from people, but all the ones i have tried have been poor compared to what some get, others have said their's are poor as well. A flagship shouldnt be so inconsistent between devices.
So ran androbench, the s7 almost never even ranked in top 10
A couple of times was like 9. The note 5 on 6 occasions was top 3 and almost all top 10. So maybe bad memory controllers on 3 repetitive s7s?
strokercrate said:
I did run geekbench but it doesn't show i/o, everyone that shows i/o shows a considerable less on s7, quadrant is another that does and antutu, and those two show ram as better in the note 5.
Never ran storage on s7, so will now.
What variant do you have?
The s7 seems to have the largest variation in performance between multiple devices i have seen based on response from people, but all the ones i have tried have been poor compared to what some get, others have said their's are poor as well. A flagship shouldnt be so inconsistent between devices.
So ran androbench, the s7 almost never even ranked in top 10
A couple of times was like 9. The note 5 on 6 occasions was top 3 and almost all top 10. So maybe bad memory controllers on 3 repetitive s7s?
Click to expand...
Click to collapse
I have the exynos version so that could explain a lot, but I wouldn't really trust benchmarks and just go by what you know and apparently the software is way worse on the snapdragon version as it stands right now (gear VR performs worse, worse battery, overheating, etc.). Just wait a couple of iterations and the problem should sort itself out. I always go for the exynos version as I've always felt that the snapdragon versions have performed worse (ever since the s4). Even bought the Note 4 in Dubai in order to be the n910c which even by today's standards performs admirably.
Well, i completely agree, i absolutely hate snap dragon versions and was upset when they switched back from the exynos in previous gen. I never felt it was as good even back in the gs3. Only benefit then was lte, but now thats not an issue i wish they would quit Qualcomm all together.
Being on cdma i had no choice this time around.
My Gear VR absolutely sucks with my S7. This is absolutely pathetic!!! When I first got my Gear VR innovator edition with my Note 4 it was probably the best gadget I had ever used, but then... Verizon/Samsung released an update that totally ruined the experience. It made the VR heat up in just a couple minutes and was basically unusable. MANY long months later they released an update to finally fix it.
I was super excited to my S7 and the new gear VR, and yet I'm right back in the same crappy boat. Not only does the performance suck, but the fit is much worse on the consumer edition and view is obstructed by the edges of the VR, and light is able to penetrate on the sides. It is so laggy/choppy/slow with my S7 that it is almost unusable. I tried it with my wife's S6 Edge + and it is absolutely perfect! I am so pissed off with Samsung that I'm ready to make the switch to the dark side!
As much as it pains me to say this, this is why Samsung will never catch Apple. Apple would never let something like this happen and even if it did happen, they would be all over it to remedy the issue. One thing that is even worse is that this was a time for Samsung to really show people how great their products are yet again they've screwed the pooch.
Sorry for the rant and not meaning to turn this into an Apple vs Samsung/Android debate, but I am really pissed that I still have to deal with this crap with this "latest and greatest" S7 Edge and Gear VR and yet it ends up being an inferior experience!!!!!!![emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853]
Sent from my iPad using Tapatalk
z06mike said:
My Gear VR absolutely sucks with my S7. This is absolutely pathetic!!! When I first got my Gear VR innovator edition with my Note 4 it was probably the best gadget I had ever used, but then... Verizon/Samsung released an update that totally ruined the experience. It made the VR heat up in just a couple minutes and was basically unusable. MANY long months later they released an update to finally fix it.
I was super excited to my S7 and the new gear VR, and yet I'm right back in the same crappy boat. Not only does the performance suck, but the fit is much worse on the consumer edition and view is obstructed by the edges of the VR, and light is able to penetrate on the sides. It is so laggy/choppy/slow with my S7 that it is almost unusable. I tried it with my wife's S6 Edge + and it is absolutely perfect! I am so pissed off with Samsung that I'm ready to make the switch to the dark side!
As much as it pains me to say this, this is why Samsung will never catch Apple. Apple would never let something like this happen and even if it did happen, they would be all over it to remedy the issue. One thing that is even worse is that this was a time for Samsung to really show people how great their products are yet again they've screwed the pooch.
Sorry for the rant and not meaning to turn this into an Apple vs Samsung/Android debate, but I am really pissed that I still have to deal with this crap with this "latest and greatest" S7 Edge and Gear VR and yet it ends up being an inferior experience!!!!!!![emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji35][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853][emoji853]
Sent from my iPad using Tapatalk
Click to expand...
Click to collapse
Smooth as butter here. Are you using the snapdragon version? Have you tried a complete reset?
Just read it.
https://www.anandtech.com/show/12520/the-galaxy-s9-review/5
https://www.anandtech.com/show/12520/the-galaxy-s9-review/8
I really hope Samsung fixes this "issues".
It sucks. Hard.
What is the problems with the exynos processor this year?
You think software issue or harware issue?
Whats wrong with Exynos cpu. Well it is the tortoise and the hare.
Intel speed shift it is not.
htcplussony said:
Whats wrong with Exynos cpu. Well it is the tortoise and the hare.
Intel speed shift it is not.
Click to expand...
Click to collapse
Exynos 9810 seems to be another 5410: made to impress in synthetic benchmarks, and stop at nothing to achieve that.
I will take this review with a grain of salt....agreed most reviewers don't go so deep and basically review their devices after a week of daily usage....but still if the problems were so deep..it would have been reported.
Like i said in the comments there, i believe most of the deductions from that test will not impact your day to day life. Information is great, but one must know when you need a "summary" and when a "thesis"
I am still buying this beauty next month and i am sure i would be satisfied.
Ive heard reviewers say and show on video the exynos having better battery life and a faster processor. Who to believe? I returned my snapdragon for an exynos because my SD got such horrible battery life. Just got the exynos so we will see. At least it can have root.
Sachinfan said:
I will take this review with a grain of salt....agreed most reviewers don't go so deep and basically review their devices after a week of daily usage....but still if the problems were so deep..it would have been reported.
Like i said in the comments there, i believe most of the deductions from that test will not impact your day to day life. Information is great, but one must know when you need a "summary" and when a "thesis"
I am still buying this beauty next month and i am sure i would be satisfied.
Click to expand...
Click to collapse
Grain of salt? Did you even read the review?!
Sure, Exynos this year will still feel smooth, just like it did last year; but there is no denying to the fact that Samsung is still not up in its game when it comes to GPU efficiency. Samsung will still need to tune the scheduler on this Exynos (if not already tuned with March update). And unlike those most phone reviews out there, Anandtech folks actually go back and update their reviews after substantial changes that manufacture might have made (AT also reports those suggestions to OEMs when they review).
Anandtech is one of the few tech sites remaining that review hardware like how hardware should be reviewed.
Do not dismiss such hard work just for the sake of sounding edgy on the internet.
Sachinfan said:
I will take this review with a grain of salt....agreed most reviewers don't go so deep and basically review their devices after a week of daily usage....but still if the problems were so deep..it would have been reported.
Like i said in the comments there, i believe most of the deductions from that test will not impact your day to day life. Information is great, but one must know when you need a "summary" and when a "thesis"
I am still buying this beauty next month and i am sure i would be satisfied.
Click to expand...
Click to collapse
I had the s9 for 10 days, sold it because I got much worse battery life than on my nearly one year old s8 and there were nearly no improvements at all. I only noticed these 3:
1. fp reader, better placement and much more accurate.
2. Sound from speakers which i never use, much better.
3. dual-sim.
In the end i felt that selling the s8 for 400 dollars more than I would get for my s8 was definitely worth it. Very happy with that decision.
As i said , such reviews have no real benefit for me as a everyday consumer. There's nothing to say that the author hasn't done an excellent job...he has. Point is whether you need such deep analysis , when you are never going to be affected. My view is <1% will notice these things.
Secondly battery is a very subjective thing, it depends on user to user and companies generally take a month or two to fix their stuff after a new release.
So I've had the exynos s9 all day and I gotta say it's much better battery life than my snapdragon s9. Only problem is now LTE doesn't work. Ughhh.
Even with a Exynos cpu, it will has godly performance.
Skickat från min SM-G960F via Tapatalk
madnav said:
Grain of salt? Did you even read the review?!
Sure, Exynos this year will still feel smooth, just like it did last year; but there is no denying to the fact that Samsung is still not up in its game when it comes to GPU efficiency. Samsung will still need to tune the scheduler on this Exynos (if not already tuned with March update). And unlike those most phone reviews out there, Anandtech folks actually go back and update their reviews after substantial changes that manufacture might have made (AT also reports those suggestions to OEMs when they review).
Anandtech is one of the few tech sites remaining that review hardware like how hardware should be reviewed.
Do not dismiss such hard work just for the sake of sounding edgy on the internet.
Click to expand...
Click to collapse
Yes it's in depth but not always how it really is.
Look at his benchmark charts and realize that SD835 is almost every benchmark faster than Exynos 8895 and 9810. Well, we all know it's not the case in real life.
Apple's A11 is fastest soc in every benchmark but in real world iPhone 8 is nowhere faster than Android flagships depending on sort of apps you use.
Never trust benchmarks only. User experience is more important.
garret1976 said:
Yes it's in depth but not always how it really is.
Look at his benchmark charts and realize that SD835 is almost every benchmark faster than Exynos 8895 and 9810. Well, we all know it's not the case in real life.
Apple's A11 is fastest soc in every benchmark but in real world iPhone 8 is nowhere faster than Android flagships depending on sort of apps you use.
Never trust benchmarks only. User experience is more important.
Click to expand...
Click to collapse
Those are not benchmarks; those are to test specific aspects of a SOC.
SD835 does still beat Exynos 8895 and 9810 when it comes to GPU intensive tasks; that is to credit QComm's exceptionally mature Adreno architecture.
SD835 will continue to beat these 2 CPU's when it comes to performance per Watt due to that.
SD810 was a failure, SD820 managed to narrow the GAP with E8890, and SD835 more or less went head to head with E8895.
Sure, synthetic benchmarks still favoured Exynos year after year, but the reality has not been that clear due to Samsung's DVFS being a year late compared to QComm. Those things might not be relevant to how a device performs (due to difference in choice of system variables and tuning); but they are very much relevant to judging how those SOC were designed and implemented. That is what the point of those tests are (over AnandTech and for their usual style of reviewing stuff).
You can check speedtests on ytube for SD845 vs E9810 and decide for yourself (speed tests are usually dumb across different devices but you can make some relevance when it comes to same device with different chips).
One such example:
AnandTech has published a testing article here for those of you who have been following this discussion and/or interested in improvements that Samsung and/or developers can bring to Exynos S9/S9+ via kernel modifications.
Improving The Exynos 9810 Galaxy S9: Part 1
Interesting read.
TL/DR:
Today’s results are just meant as a quick preview of the Exynos S9’s behaviour when removing what I believed the most offending parts of Samsung’s mechanisms. The results are very promising as we are essentially getting real-world performance boosts while at the same time seeing efficiency improvements and battery life improvements.
There is still a lot to be done - I haven’t really touched the scheduler or DVFS scaling logic, but I’m confident that it’s possible to improve things further. It’s unlikely that we’ll end up matching the Snapdragon 845 variant in battery life, as there are efficiency curve considerations at lower performance points that simply cannot be changed through software. But closing the gap as much as possible seems to be an attainable short-term goal.
Click to expand...
Click to collapse
Now Samsung chimed in on the issue
https://www.androidcentral.com/samsung-addresses-battery-issues-exynos-powered-galaxy-s9
I'm extremely intrigued but also confused at the same level. I ordered a regular S9 Dual-SIM after reading reviews and reports that told me it was a great device and a nice upgrade from my S7 Edge. Then I cancelled my order because I read that a lot of people with the Exynos variant have battery related issues. Then I kept reading and noticed most of these reports where from devices just a couple hours or days old. The more recent reviews seem to show a different picture. While still a mixed bag, battery performance seems to be much less of an issue in more recent reports. For some it's even better than their previous devices. The reviews/reports who seem to praise the performance (battery) the most include almost exclusively people who compare it to the S6 (Edge), S7 (Edge). So, I placed my order again for a Coral Blue S9 Dual-SIM. Don't care for the Dual-SIM but somehow it's the only variant available in Germany. Whatever.
This is by no means an accurate analysis of the current or past situation. It's my experience. It's not representative in any form or function.
nitrous² said:
I'm extremely intrigued but also confused at the same level. I ordered a regular S9 Dual-SIM after reading reviews and reports that told me it was a great device and a nice upgrade from my S7 Edge. Then I cancelled my order because I read that a lot of people with the Exynos variant have battery related issues. Then I kept reading and noticed most of these reports where from devices just a couple hours or days old. The more recent reviews seem to show a different picture. While still a mixed bag, battery performance seems to be much less of an issue in more recent reports. For some it's even better than their previous devices. The reviews/reports who seem to praise the performance (battery) the most include almost exclusively people who compare it to the S6 (Edge), S7 (Edge). So, I placed my order again for a Coral Blue S9 Dual-SIM. Don't care for the Dual-SIM but somehow it's the only variant available in Germany. Whatever.
This is by no means an accurate analysis of the current or past situation. It's my experience. It's not representative in any form or function.
Click to expand...
Click to collapse
The upgrade will be very notable over s7 generation.
S7Edge gave me great battery (6-7hr SOT); S8 and S8+ gave me lower than that (4-5hr SOT). Now S9+ again gives me great battery (6-7hr SOT).
All of them have been Dual SIM models; and yes, I did change phone that often since I had a way to upgrading for a very small price difference.
S9+ upgrade wasn't as cheap for me, so I'll have to continue on this for a bit longer. It is an exynos S9+ from India; and performance has not been an issue so far for me.
Sent from my SM-G965F using Tapatalk
I had the regular s7 before and upgraded to the exynos variant s9, and it is true that battery life during the first week or two is not adequate. Battery life has improved much and overall it's still a great device compared to other devices out there. Yes, I do wish the battery capacity is larger and Samsung do further optimizations to get battery life on par with the snapdragon variant it isn't the worse and I'm able to get a full day out of it. Honestly if you can wait, wait for the s10. The new 7nm chips and rumored higher battery capacity should make a killer device. But you can't go wrong with the s9 believe me. Other devices out there are flawed one way or another and the touchwiz experience is honestly approaching perfection, yes it was terrible years back but now it's way better than stock android in my opinion.
In the screenshot below, the power saving mode only has decreased 10% screen brightness only. No other altered settings.
stevenl23 said:
I had the regular s7 before and upgraded to the exynos variant s9, and it is true that battery life during the first week or two is not adequate. Battery life has improved much and overall it's still a great device compared to other devices out there. Yes, I do wish the battery capacity is larger and Samsung do further optimizations to get battery life on par with the snapdragon variant it isn't the worse and I'm able to get a full day out of it. Honestly if you can wait, wait for the s10. The new 7nm chips and rumored higher battery capacity should make a killer device. But you can't go wrong with the s9 believe me. Other devices out there are flawed one way or another and the touchwiz experience is honestly approaching perfection, yes it was terrible years back but now it's way better than stock android in my opinion.
In the screenshot below, the power saving mode only has decreased 10% screen brightness only. No other altered settings.
Click to expand...
Click to collapse
This is what I'm talking about.
Battery tests, even in-depth semi-scientific analyses are all over the place and inconclusive. It is true that battery life is very user dependant but there should be a common theme all across the board. However, I don't see anything like that. I can't dealt with this for 2 years. I cancelled my order and will wait for a better deal on the S9+.
So , it seems , my reaction to the anandtech review was spot-on. S9 , irrespective of the cpu variants, is still a great device and we must not confuse ourselves with too much information where none is reqd.
God willing, I should be getting my hands on the s9 in early May.
Very interesting read!
https://www.anandtech.com/show/12520/the-galaxy-s9-review
If Samsung doesn't fix this I'll sell my S9
Really wanted to buy one but I'm not going to as long as this issue is not fixed.
Fix what? It looks like any performance "fix" will be accompanied by loss of battery life anyway. There is a slim possibility that scheduler/governor could be improved somewhat, but that will not change the underlying hardware characteristics that much, meaning the gain from such a fix will not be significant and you will still pay a hefty battery cost..
Like i said in the comments there, i believe most of the deductions from that test will not impact your day to day life. Information is great, but one must know when you need a "summary" and when a "thesis"
I am still buying this beauty next month and i am sure i would be satisfied.
lopri said:
Fix what? It looks like any performance "fix" will be accompanied by loss of battery life anyway. There is a slim possibility that scheduler/governor could be improved somewhat, but that will not change the underlying hardware characteristics that much, meaning the gain from such a fix will not be significant and you will still pay a hefty battery cost..
Click to expand...
Click to collapse
Whenever a load is transferred from LITTLE cores to the big cores, SD845's big cores start at the max frequency while E9810's big cores start at the lowest frequency (and then increase gradually if load continues to be active).
If the load is of a longer nature (beyond 450ms iirc) then E9810 wins the race; but for short and burst loads the SD845 wins the race due to starting at max frequency.
So no, scheduler fix will not improve just somewhat; it will improve the performance substantially.
Sure, that might come with a battery hit since it will be a choice between 'a longer running load at lower power' vs 'a shorter running load at high power'; we will know better about battery drains when we get results from having used such a scheduler.
The way that exynos takes that long to ramp up the cpu, its like the only way one ever gets to 'experience' it in full speeds is synthetic benchmarks.
I was pumped coming from Snapdragon devices and thought the general consensus was that Exynos chips were typically better than Snapdragon. I'm saving to buy the Exynos version, but then a few of the big Youtubers ran speedtests of SD845 and 9810 and it seemed like the 845 generally performed better.
And after reading the anadtech article and battery life being worse (which is huge to me because I want the regular S9 and not the plus) I guess Snapdragon all the way.
Nurmi_CEO said:
I was pumped coming from Snapdragon devices and thought the general consensus was that Exynos chips were typically better than Snapdragon. I'm saving to buy the Exynos version, but then a few of the big Youtubers ran speedtests of SD845 and 9810 and it seemed like the 845 generally performed better.
And after reading the anadtech article and battery life being worse (which is huge to me because I want the regular S9 and not the plus) I guess Snapdragon all the way.
Click to expand...
Click to collapse
I think general every day battery life depends on a lot more than just benchmarking. I have both the exynos and snapdragon S9 with me right now and the exynos gets better battery life after 1 day of use. I just got the phone yesterday but both starting with a 100% charge and using both pretty equal amounts and doing testing and web surfing etc... my exynos has about 10% more right now. Driving to work yesterday my SD845 phone drained 14% with the screen off and not even music playing. Honestly coming from a Pixel 2 XL they both kinda suck hard on battery life. The SD version was dying in half a day of light use for me. I got through the rest of the day yesterday with the exynos which is doable but ill have to keep chargers all around to make sure it stays charged on heavy use days.
Edit: On a side note if you decide to get the exynos and you use T-Mobile you may want to wait because I get horrible reception and slow internet speeds. I dont know why sicne it supports every band for LTE except 71 but the snapdragon version is significantly better and much faster on tmobile network. Maybe there are settings I can change. IDK.
---------- Post added at 06:56 PM ---------- Previous post was at 06:53 PM ----------
On a side note if you decide to get the exynos and you use T-Mobile you may want to wait because I get horrible reception and slow internet speeds. I dont know why sicne it supports every band for LTE except 71 but the snapdragon version is significantly better and much faster on tmobile network. Maybe there are settings I can change. IDK.
I would really like to have Band 71. I thought I saw a post from someone who was getting conflicting answers from Samsung if B71 was in the international unlocked version. However then he posted and said he received it and sure enough found B71 in it. But now it's looking like I'll just bite the bullet and get the T-Mo SD845 version. Would have be nice to have an unlocked bootloader.
Nurmi_CEO said:
I would really like to have Band 71. I thought I saw a post from someone who was getting conflicting answers from Samsung if B71 was in the international unlocked version. However then he posted and said he received it and sure enough found B71 in it. But now it's looking like I'll just bite the bullet and get the T-Mo SD845 version. Would have be nice to have an unlocked bootloader.
Click to expand...
Click to collapse
How can you find out if it has Band 71 on it? I can test side by side my US and international verison and the US version gets about 30-40mbps and right beside it is the international getting about 1.5-3mbps max. Soemthing is obviously wrong with the settings or the radio on the international to be doing that bad on tmobile. Its not bad location. Something is just missing either software, hardware or settings but I just odnt know. Id love to make this thing work.
jrharvey said:
I think general every day battery life depends on a lot more than just benchmarking. I have both the exynos and snapdragon S9 with me right now and the exynos gets better battery life after 1 day of use. I just got the phone yesterday but both starting with a 100% charge and using both pretty equal amounts and doing testing and web surfing etc... my exynos has about 10% more right now. Driving to work yesterday my SD845 phone drained 14% with the screen off and not even music playing. Honestly coming from a Pixel 2 XL they both kinda suck hard on battery life. The SD version was dying in half a day of light use for me. I got through the rest of the day yesterday with the exynos which is doable but ill have to keep chargers all around to make sure it stays charged on heavy use days.
Edit: On a side note if you decide to get the exynos and you use T-Mobile you may want to wait because I get horrible reception and slow internet speeds. I dont know why sicne it supports every band for LTE except 71 but the snapdragon version is significantly better and much faster on tmobile network. Maybe there are settings I can change. IDK.
---------- Post added at 06:56 PM ---------- Previous post was at 06:53 PM ----------
On a side note if you decide to get the exynos and you use T-Mobile you may want to wait because I get horrible reception and slow internet speeds. I dont know why sicne it supports every band for LTE except 71 but the snapdragon version is significantly better and much faster on tmobile network. Maybe there are settings I can change. IDK.
Click to expand...
Click to collapse
jrharvey said:
How can you find out if it has Band 71 on it? I can test side by side my US and international verison and the US version gets about 30-40mbps and right beside it is the international getting about 1.5-3mbps max. Soemthing is obviously wrong with the settings or the radio on the international to be doing that bad on tmobile. Its not bad location. Something is just missing either software, hardware or settings but I just odnt know. Id love to make this thing work.
Click to expand...
Click to collapse
You'll have to look through the forum here. I have other things going on, but I remember someone created a thread and stated that it did. I have no further info.
Well, it's not like the phone is slow in any way so stop making a chicken of a feather.
Skickat från min SM-G960F via Tapatalk
jrharvey said:
I can test side by side my US and international verison and the US version gets about 30-40mbps and right beside it is the international getting about 1.5-3mbps max.
Click to expand...
Click to collapse
I've been holding out on an upgrade for the S9, but this is concerning (I use T-Mobile).
Snapdragon is a non-option for me as having proper backup (aka Titanium) is IMO a bare essential. Now after having seen reviews that the S9 Exynos' battery life is even worse than the S8's, I'm wondering if I should just get the S8 instead. Anyone have any thoughts - S8 Exynos or S9 Exynos? If that Anandtech article's results are at all representative of what to expect, does it not seem silly to pay more for a newer phone that's a *downgrade* in battery life? Or maybe I just need to wait a bit longer for the kinks to get worked out? I mean, if it can really only do 1.5mbps on T-Mobile...ummm....
metal450 said:
I've been holding out on an upgrade for the S9, but this is concerning (I use T-Mobile).
Snapdragon is a non-option for me as having proper backup (aka Titanium) is IMO a bare essential. Now after having seen reviews that the S9 Exynos' battery life is even worse than the S8's, I'm wondering if I should just get the S8 instead. Anyone have any thoughts - S8 Exynos or S9 Exynos? If that Anandtech article's results are at all representative of what to expect, does it not seem silly to pay more for a newer phone that's a *downgrade* in battery life? Or maybe I just need to wait a bit longer for the kinks to get worked out? I mean, if it can really only do 1.5mbps on T-Mobile...ummm....
Click to expand...
Click to collapse
Ok well I did so many things to try and fix this and SOMETHING worked. Im now getting pretty much full speed although still not quite as good as teh US version for some reason although very very close. I went to Tmobile and got a new sim card in case this one was damaged or soemthing. i also played around with the APN settings and restarted several times. Something jolted the phone into working. SO when I test the two phones side by side now the US version might get 21mbps and the international 18mbps. Again the US will get 31mbps and the international 26mbps. US 13mbps and international 11mbps. The US version is always just slightly ahead. I drove around town and found some low reception spots where the US version was getting 2mbps and the international was crawling at .65 mbps. Out of probably 30 different test I did driving around I think only 1 time the international got better speeds and it was like 31mbps vs the US getting like 28. Id say its usable for sure but here is where it might matter most. I was in a bad reception part of town in a big mall type building and I rarely ever get service there. You have to stand next to a window to get any bars at all. The US version could barely get a connection but it did get something. The international didnt see anything until I completely left the building. Coming from a google pixel 2 XL that phone was about the same as what the US version of the galaxy s9 got. I walked all around the building and saw similar results. We dont even have band 71 in my city and that is the ONLY band that is missing from the international version that I have. Theoretically the international version SHOULD be performing the same however ive done plenty of test I can pretty much say in my case the international is slower.
Ironically the International version has been killing the US version in battery life. Im very happy about that. In standby mode the US version seems to drain very quickly for me while the exynos seems to just barely sip power. I can go an hour in standby and only use 1-2% while the US version can easily use 8% just sitting there and not doing anything. This morning I woke up and cleaned the house and both had 100% when I started. I didnt touch the phones at all. Both on wifi. About an hour later the exynos was at 98% and the US was at 93%. Both phones have the same apps and theoretically the same services running. The international actually SHOULD have more running in the background since my allo accound and texts/calls go to that phone and my wifes sim is in the other phone.
jrharvey said:
Im now getting pretty much full speed although still not quite as good as teh US version for some reason although very very close.
Click to expand...
Click to collapse
Well that's relieving to hear. 1 of my 3 concerns ticked off the list!
jrharvey said:
I was in a bad reception part of town in a big mall type building and I rarely ever get service there. You have to stand next to a window to get any bars at all. The US version could barely get a connection but it did get something. The international didnt see anything until I completely left the building.
Click to expand...
Click to collapse
Damn, that's definitely an issue, especially since I live in a very low-signal part of town (in the mountains). I wonder how Exynos S8 vs Exynos S9 does in this regard...or if anyone else has any similar experiences as you...
jrharvey said:
Ironically the International version has been killing the US version in battery life.
Click to expand...
Click to collapse
Super odd, considering that's the exact opposite of what all the reviews seem to be saying! I'd be curious if anyone else could replicate this result - or again, to see something similar with Exynos S8 vs Exynos S9...
One more question: have you been able to get T-Mobile Wifi calling on the Exynos S9? I'm guessing not, but here in my low-signal area that's pretty essential. I figure someone on here will eventually work it out, but I was just hoping it might somehow already be up & running
metal450 said:
Well that's relieving to hear. 1 of my 3 concerns ticked off the list!
Damn, that's definitely an issue, especially since I live in a very low-signal part of town (in the mountains). I wonder how Exynos S8 vs Exynos S9 does in this regard...or if anyone else has any similar experiences as you...
Super odd, considering that's the exact opposite of what all the reviews seem to be saying! I'd be curious if anyone else could replicate this result - or again, to see something similar with Exynos S8 vs Exynos S9...
One more question: have you been able to get T-Mobile Wifi calling on the Exynos S9? I'm guessing not, but here in my low-signal area that's pretty essential. I figure someone on here will eventually work it out, but I was just hoping it might somehow already be up & running
Click to expand...
Click to collapse
I agree the battery life thing is very odd when you look at the word around town about the exynos. I think all that talk is sparked by the Adandtech review of the two phones and what you have to consider is those are synthetic benchmarks and they were also dealing with the snapdragon S9+ and the exynos S9 so completely different screen sizes and battery sizes. Now they used the cpu efficiency to come up with this idea that the exynos would theoretically perform worse than the snapdragon but thats all synthetics. I saw many online reviews of the battery life from the exynos and snapdragon versions and compared them and took a leap of faith that the exynos would be better. That was before the anandtech review came out. Thank god I was kinda right because battery life could not have been worse than my snapdragon s9 which just seemed to kill battery even when screen off and idle. Maybe its just my unit was bad but who knows. My review of battery life is strictly based on normal every day use and not synthetic benchmarks that nobody just does all day every day. Benchmarks are nice but they dont do crap for how you actually use the phone.
jrharvey said:
I agree the battery life thing is very odd when you look at the word around town about the exynos. I think all that talk is sparked by the Adandtech review of the two phones and what you have to consider is those are synthetic benchmarks and they were also dealing with the snapdragon S9+ and the exynos S9 so completely different screen sizes and battery sizes. Now they used the cpu efficiency to come up with this idea that the exynos would theoretically perform worse than the snapdragon but thats all synthetics.
Click to expand...
Click to collapse
For me that particular comparison is irrelevant, because a Snapdragon is non-unlockable -> non-backupable -> non-option. I was concerned about the results showing Exynos S9 is even inferior to Exynos S8.
And I agree that every-day use is what matters, not synthetics. However, I would also assume that synthetics would have *some* correlation to real-world; obviously they aren't a perfect representation of it, or they may exaggerate the difference, etc. But it's pretty surprising to find the exact opposite result. Makes me wonder if the S8 vs S9 result could be backwards too...
Give me an exynos over a snapdragon anyday.
Im more than happy with a small loss of performance over not being able to unlock my bootloader and customize to my hearts content.
My s9 plus is a very good upgrade on my s8 (now with the wife)
Im not even going to read this article as its a moot point tbh. Real world performance wise, this phone blows anything else i have used out of the water.
Yes with both battery and performance abd despite it being an exynos lol.
Performance is excellent with my galaxy s9 it's just battery I wish was better.
I used my wife's Note 9 (Snapdragon - ATT ) and ran ANTUTU Benchmarks for grins and giggles and the phone got a score of about 284,000 - very respectable.
When I received my N960F - a Note 9 from Clove in the UK, I ran the same Benchmark from ANTUTU and the best score I could get was 236,000 - a HUGE difference.
I really expected more parity between those phones and wonder why there is such a big difference? Is the Exynos Process that far behond Snapdragon's latest?
(I know these benchmarks don't mean much - but just the same, after spending nearly a grand on this phone, I would have expected better)
Any thoughts?
same chips but on the S9. should give you an idea
Geekser said:
I used my wife's Note 9 (Snapdragon - ATT ) and ran ANTUTU Benchmarks for grins and giggles and the phone got a score of about 284,000 - very respectable.
When I received my N960F - a Note 9 from Clove in the UK, I ran the same Benchmark from ANTUTU and the best score I could get was 236,000 - a HUGE difference.
I really expected more parity between those phones and wonder why there is such a big difference? Is the Exynos Process that far behond Snapdragon's latest?
(I know these benchmarks don't mean much - but just the same, after spending nearly a grand on this phone, I would have expected better)
Any thoughts?
Click to expand...
Click to collapse
What are your thoughts on battery life between the 2?
lawtq said:
What are your thoughts on battery life between the 2?
Click to expand...
Click to collapse
great question but I haven't used either long enough to really get a sense for it - On first blush, battery seems similar on both - I can easliy get thru an entire day without charging - though I find myself automatically placing my phone in my wireless charger from time to time because I have always had to - I think I can break myself of that habit though - will play around a bit more on both and see how they fair in a comparison of daily use....
soraxd said:
same chips but on the S9. should give you an idea
Click to expand...
Click to collapse
lawtq said:
What are your thoughts on battery life between the 2?
Click to expand...
Click to collapse
Geekser said:
great question but I haven't used either long enough to really get a sense for it - On first blush, battery seems similar on both - I can easliy get thru an entire day without charging - though I find myself automatically placing my phone in my wireless charger from time to time because I have always had to - I think I can break myself of that habit though - will play around a bit more on both and see how they fair in a comparison of daily use....
Click to expand...
Click to collapse
Did Samsung work on the kernel sources to improve the Exynos 9810's efficiency and performance to be on par with Snapdragon 845 ?
The answer is "NO".
Looks like the the tides have turned recently in favor of the snapdragon chipset.
Sent from my SM-N960U1 using Tapatalk
Virgo_Guy said:
Did Samsung work on the kernel sources to improve the Exynos 9810's efficiency and performance to be on par with Snapdragon 845 ?
The answer is "NO".
Click to expand...
Click to collapse
I know and I don't understand why. I think Exynos could be as good or better than the Snapdragon - but for some reason - it lags behind...
Geekser said:
I know and I don't understand why. I think Exynos could be as good or better than the Snapdragon - but for some reason - it lags behind...
Click to expand...
Click to collapse
Could the Exynos 9810 be better in the s9/s9+/Note 9? - Probably to an extent working on the kernel but would still be behind the SD.
Geekser said:
great question but I haven't used either long enough to really get a sense for it - On first blush, battery seems similar on both - I can easliy get thru an entire day without charging - though I find myself automatically placing my phone in my wireless charger from time to time because I have always had to - I think I can break myself of that habit though - will play around a bit more on both and see how they fair in a comparison of daily use....
Click to expand...
Click to collapse
Thanks. Look forward to hearing from you
Snapdragon Note 9 : Will Allow you to Play FASTER Gamecube/Wii/PSP Emulation. But Will not allow you to install custom roms or root your device.
Exynos Note : Will not play gamecube/Wii Games Smoothly but will allow you to install custom rom.
Personally if I didn't want Wii/Gamecube emulation, I would have easily bought exynos.
Ok in synthetic benchmarks such as antutu the 845 gets a higher score because of the better graphics processor. When it comes to the cpu side of things the 9810 is at least as good and sometimes better.
Cool Person said:
Snapdragon Note 9 : Will Allow you to Play FASTER Gamecube/Wii/PSP Emulation.But Will not allow you to install custom roms or root your device.
Exynos Note : Will not play gamecube/Wii Games Smoothly but will allow you to install custom rom.
Personally if I didn't want Wii/Gamecube emulation, I would have easily bought exynos.
Click to expand...
Click to collapse
dolphin emulation on both chipsets recently took a massive jump forward. exynos included.
you can root on international snapdragon variants of the Note 9. I believe China has unlocked snapdragon. I've heard chit chat mentioning Canada as well, fact check these yourself tho.
mikey_sk said:
Ok in synthetic benchmarks such as antutu the 845 gets a higher score because of the better graphics processor. When it comes to the cpu side of things the 9810 is at least as good and sometimes better.
Click to expand...
Click to collapse
Aha, but then the real world performance of the sd is far better than the exynos + it's more efficient leading to a lot better battery life + the modem is better. Then you throw in a GPU that is literally another class in a phone that potentially will be used also for gaming and this year a lot of heavy games were released with vastly different experience on the two variants (last year samsung got a free pass on their lossy g71 mp20 as no heavy games released, not the case this year and it's shaping more to release frequently in the future).
g72 mp18 can't even come close to the PREVIOUS year sd 835 in real world gaming. sd 845 benefits heavily from the note 9 cooling as it had some thermal restrictions and sustain performance was close to the sd 835 one (still far better than g72 mp18 sustain for both snapdragons), while the mali GPU doesn't benefit that much from the new cooling, the GPU is just slow.
In Europe we actually pay more for that phone, 1000 euros are not equal to 1000 us dollars, all that while we get a SOC that is between midrange and highend instead of the highend one that is for the US market. Previous years the exynos was better (galaxy s7 for example), but the difference between the two variants were never that big as this year.
Edit: check the other real world benchmarks like pcmark, browser benchmarks and so on - the sd variant wipes the floor with the exynos. Still, benches are not all, sadly the real world difference is also there and noticeable.
Again if you look at the geek bench scores for cpu (android central, did a comparison) the cpu side of the exynos 9810 outperformed the 845. I would agree when factoring in the 845 graphics the overall score of the 845 is better. Again you would be VERY hard pressed to notice the difference in the real world. And there is zero lag on either of these variants I found.
I know people in the industry and alot of the perceived difference is due to the scheduler for the 9810 not being as optimised as for the 845. Rule on the street is that Sammy will redress this soon with an update.
Also the woolfson DAC on the exynos provides a better quality audio experience and fuller sound stage than that of the aqstic DAC on the 845.
Again the difference is slim. I've noticed zero lag even on fortnite.
Reast assured both chips ate brutes. After all both the 845 and 9810 are currently fabricated/manufactured by samsung.
mikey_sk said:
Again if you look at the geek bench scores for cpu (android central, did a comparison) the cpu side of the exynos 9810 outperformed the 845. I would agree when factoring in the 845 graphics the overall score of the 845 is better. Again you would be VERY hard pressed to notice the difference in the real world. And there is zero lag on either of these variants I found.
I know people in the industry and alot of the perceived difference is due to the scheduler for the 9810 not being as optimised as for the 845. Rule on the street is that Sammy will redress this soon with an update.
Also the woolfson DAC on the exynos provides a better quality audio experience and fuller sound stage than that of the aqstic DAC.
Again the difference is slim. I've noticed zero lag even on fortnite.
Reast assured both chips ate brutes. After all both the 845 and 9810 are currently fabricated/manufactured by samsung.
Click to expand...
Click to collapse
It's a lot more complicated than a simple scheduler and if it was going to be fixed, it was going to be fixed for the note 9 launch as it's a good few months from the s9/s9+ release with the same SOC (that is a lot in the fast moving forward mobile industry). Most likely they won't do any improvements even in future major android releases. Geekbench score is peak performance on burst load tasks, it's not counting task migrations, frequency scaling, cores hotpluging and so on. There are good articles for the 9810 in anandtech, how it can be improved, why it can not be improved beyond given point, but they are kinda techy and not easy to read if you are not into it.
tl:dr The CPU cores are great, but has some flaws and everything around them is sh*t (sorry for my language, but yeah). The DAC is great tho, this is totally true.
For the GPU part there are a lot of materials about adreno 630 vs g72 mp18 with actual stats like max fps, min fps, FPS stability and so on. Fortnite will run ok, but on the sd845 it can run maxed out fluently (for the most part). Also when one spend 1000 euro, he should think about future proving and gaming is moving fast forward this year. If you game on your phone, it's kinda a bummer to settle for lower GFX settings on one of the most expensive phones this year, if you want smooth experience.
Not saying the exynos is super bad, it's not, but it's not on par with the sd845.
Again you would be hard pressed to notice the difference. Both variants are perfectly fine so everyone buy either with confidence.
It's not true, marvel strike force is unplayable on exynos version
9810 user here and i have to limit the exynos speed for it to get to a day. This sucks. I paid usd1000 but need to turn pn battery saver every morning
No rows written here reflect the reality. Just snapdragon users happy with their chipset. Another report on the S9 test that is outdated. The smallest SOT with exynos is 7h with only mobile data. Who does not know how to use an android or uses it as a playstation is his job. For a normal user it's great.
roticanai said:
9810 user here and i have to limit the exynos speed for it to get to a day. This sucks. I paid usd1000 but need to turn pn battery saver every morning
Click to expand...
Click to collapse
I don't even have to do that with my Note 8. You can likely squeeze more out of it, but a lot depends on usage patterns of course.
The early reviews seem to be mixed. Has anyone managed to source two together to give a verdict what is the latest and greatest amongst the two in all aspects not just performance.
CPU seems similar, GPU better on the SD. Exynos won't be similar on GPU until it gets AMD GPU next year.
Batfink33 said:
CPU seems similar, GPU better on the SD. Exynos won't be similar on GPU until it gets AMD GPU next year.
Click to expand...
Click to collapse
Once it gets the AMD GPU it will destroy the Snapdragon and possibly finally level up with the Apple. Maybe not CPU wise, but definitely on the GPU front against Apple.
Edit. Wrong thread. Apologies.
mtm1401 said:
Click to expand...
Click to collapse
This man is not accurate because the test he ran only on one device by his own and he let another person to perform same test on that person's device then he compared both even though the one with Exynos 2100 was set to QHD and the one with SD888 was set to FHD only!! I mean this could impact the tests negatively!! The test conditions should be exactly the same!!
Listen to this guy, he is from HK and he've got the Snapdragon 888, in this video at 3:38 he mention that he don't understand how the american youtubers claims a 1.5 day battery life while his own experience in much less than that (almost half of that) .. so many youtubers unfortunately lies about the battery life and obviously this fool people who buys that device with Snapdragon 888, many of them put the phone only on Wi-Fi and not on cellular, it's not good and not nice when they claim something which is so badly incorrect. Note that BEN from the video below is using the phone with a SIM card, not with wi-fi only!
Review for SD888
nightoooo said:
This man is not accurate because the test he ran only on one device by his own and he let another person to perform same test on that person's device then he compared both even though the one with Exynos 2100 was set to QHD and the one with SD888 was set to FHD only!! I mean this could impact the tests negatively!! The test conditions should be exactly the same!!
Click to expand...
Click to collapse
Where is the source for your info?
Invincible29 said:
Where is the source for your info?
Click to expand...
Click to collapse
So in all fairness, I did go back and look at the speed test video he did between the two and noticed a few things
Looks like different people kicking off the tests (based on the table they are on and skin colour).
SD is in aeroplane mode and Exynos is connected to cellular but no wifi/mobile data.
Resolutions are very slightly different, but both look like they are in FHD, but the SD has a slightly lower resolution as the navigation bar is different and makes the screen slightly smaller. QHD makes a minimal difference to battery life unless you're doing something GPU intensive.
In all fairness, I've seen mixed reviews of both chips, but no idea how people are setting up their phones (always do a clean install!!!). Have to wait and see how it performs when the phone arrives.
mtm1401 said:
So in all fairness, I did go back and look at the speed test video he did between the two and noticed a few things
Looks like different people kicking off the tests (based on the table they are on and skin colour).
SD is in aeroplane mode and Exynos is connected to cellular but no wifi/mobile data.
Resolutions are very slightly different, but both look like they are in FHD, but the SD has a slightly lower resolution as the navigation bar is different and makes the screen slightly smaller. QHD makes a minimal difference to battery life unless you're doing something GPU intensive.
In all fairness, I've seen mixed reviews of both chips, but no idea how people are setting up their phones (always do a clean install!!!). Have to wait and see how it performs when the phone arrives.
Click to expand...
Click to collapse
Have you gone for snapdragon or exynos. I am on the fence for buying either and can't decide. Not enough reliable info but if its anything like last year then I would rather import 888 into the UK.
Invincible29 said:
Have you gone for snapdragon or exynos. I am on the fence for buying either and can't decide. Not enough reliable info but if its anything like last year then I would rather import 888 into the UK.
Click to expand...
Click to collapse
I pre-ordered the Exynos, trade in deal for my s10+ was good and I was in the market for some earbuds anyway.
I won't send my current phone in until I'm happy, 28 days should be enough to determine if the Ultra is worth keeping or not. (although won't get much of a chance to test out and about!).
I have considered SDs in the past, just a bit concerned about replacing it via insurance if it gets damaged (have to take it to a Samsung store). Seems like this year the SD isn't worth the hassle unless you constantly play demanding games which I don't. I would wait for the AMD S22, but I'm too impatient/tech whore sooooo couldn't wait.
The problem for me is trying to avoid Chinese phones (for ethical reasons), doesn't leave much choice!
mtm1401 said:
So in all fairness, I did go back and look at the speed test video he did between the two and noticed a few things
Looks like different people kicking off the tests (based on the table they are on and skin colour).
SD is in aeroplane mode and Exynos is connected to cellular but no wifi/mobile data.
Resolutions are very slightly different, but both look like they are in FHD, but the SD has a slightly lower resolution as the navigation bar is different and makes the screen slightly smaller. QHD makes a minimal difference to battery life unless you're doing something GPU intensive.
In all fairness, I've seen mixed reviews of both chips, but no idea how people are setting up their phones (always do a clean install!!!). Have to wait and see how it performs when the phone arrives.
Click to expand...
Click to collapse
Thanks a lot of sparing me time to get the info again, obviously the differences are there and i get so sick of people claiming they are doing something huge while they just needs to get more publicity than giving the pure truth to everyone!!
For my opinion, i will not accept any of the videos to let me purchase or not the S21 Ultra, because they always run the phone in "lab" conditions, not real conditions that could affect the whole experience!
Invincible29 said:
Have you gone for snapdragon or exynos. I am on the fence for buying either and can't decide. Not enough reliable info but if its anything like last year then I would rather import 888 into the UK.
Click to expand...
Click to collapse
I was in the same seat of you, not sure to get the Exynos 2100 in Sweden or i go for SD888 from HK. I have today S20 Ultra Snapdragon model and very happy with its performance & battery life but not the camera with the auto-focus issues and also for the macro pictures. So i was looking at all tests, feedbacks, reviews, i am now more sure that Exynos 2100 is for me this time because of the battery life and also the good performance, i know the GPU is weaker but for my part, battery life is more important, so with SD888 and its bad battery life there is no reason to buy it, i don't play games with my phones so i will be fine with Exynos 2100!
BTW, the Exynos 2100 gives me what i need, from full support for 5G network as my carrier has kicked off here in Sweden unlike the SD888 which lacks the important Band 40 and 42! Also the SD888 lacks the eSIM functionality!
nightoooo said:
I was in the same seat of you, not sure to get the Exynos 2100 in Sweden or i go for SD888 from HK. I have today S20 Ultra Snapdragon model and very happy with its performance & battery life but not the camera with the auto-focus issues and also for the macro pictures. So i was looking at all tests, feedbacks, reviews, i am now sure that Exynos 2100 is for me this time because of the battery life and also the good performance, i know the GPU is weaker but for my part battery life is more important, so with SD888 and bad battery time, there is no reason for me to buy it, i don't play games with my phones so i will be fine with Exynos 2100 for now.
Click to expand...
Click to collapse
I have been watching Gary's videos as well. Its a close call this time and I don't play games on the phone. Leaning towards exynos due to warranty and not having to import plus buds and tag as freebies etc. Just conscious of what happened last year and the difference was huge. Also samsung is not selling mmwave snapdragon or exynos outside of the US which is mean as mmwave is only a year or two away from landing in Europe. Why not treat all of its customers equally! I think with the recent history snapdragon had a novelty factor to it but others are catching up. Let's hope snapdragon doesn't become intel of mobile world.
nightoooo said:
Thanks a lot of sparing me time to get the info again, obviously the differences are there and i get so sick of people claiming they are doing something huge while they just needs to get more publicity than giving the pure truth to everyone!!
For my opinion, i will not accept any of the videos to let me purchase or not the S21 Ultra, because they always run the phone in "lab" conditions, not real conditions that could affect the whole experience!
Click to expand...
Click to collapse
Yea it's not exactly like for like but can't imagine it makes loads of difference. I'm also more fussed about battery life, so will test it once I get it and see how it goes, battery life is just so dependent on personal usage, although I will say Phonebuff do a better job of testing, i.e. standby times and such, which gives a better idea of relative performance.
Andrei from Anandtech does really good reviews as well, but so in-depth that they don't usually release in time for pre-order
Invincible29 said:
I have been watching Gary's videos as well. Its a close call this time and I don't play games on the phone. Leaning towards exynos due to warranty and not having to import plus buds and tag as freebies etc. Just conscious of what happened last year and the difference was huge. Also samsung is not selling mmwave snapdragon or exynos outside of the US which is mean as mmwave is only a year or two away from landing in Europe. Why not treat all of its customers equally! I think with the recent history snapdragon had a novelty factor to it but others are catching up. Let's hope snapdragon doesn't become intel of mobile world.
Click to expand...
Click to collapse
Yes, for sure you are getting 2 years warranty with Exynos 2100 in UK (i think) because we get that in Sweden too, additionally you are getting the pre booking freebies (Galaxy Buds Pro + Smart Tag), then you don't need to compare Exynos 2100 with the extremely bad Exynos 990 because the CPU is exactly the same as in SD888.
Regarding the MMwave, i don't care about it right now as it is not implemented yet in the networks here in Sweden, but i do have full 5G support for the rest of the bands which give me an advantage when the 4G is fully loaded.
I think Snapdragon will not be the intel of mobile world because Samsung is getting AMD help and involvement in the coming products, this will change the game!
This e sim is really important as you can add profiles without the need for physical sims. I had double checked this with HK version and it does not seem to be on the spec sheet. Yes band 40 will be gold once o2 pull their finger out in the UK for 5g. I think the decision is made for me that its going to be exynos for me plus also get samsung pay to work which is really handy for store cards. I'm using s9+ now and I don't know how uncomfortable the size of s21 ultra will be!
mtm1401 said:
So in all fairness, I did go back and look at the speed test video he did between the two and noticed a few things
Looks like different people kicking off the tests (based on the table they are on and skin colour).
SD is in aeroplane mode and Exynos is connected to cellular but no wifi/mobile data.
Resolutions are very slightly different, but both look like they are in FHD, but the SD has a slightly lower resolution as the navigation bar is different and makes the screen slightly smaller. QHD makes a minimal difference to battery life unless you're doing something GPU intensive.
In all fairness, I've seen mixed reviews of both chips, but no idea how people are setting up their phones (always do a clean install!!!). Have to wait and see how it performs when the phone arrives.
Click to expand...
Click to collapse
What would you consider a clean install? Can you import settings from your samsung profile? I generally don't transfer with the USB, but I download all my apps after the phone is booted up from the original install. Only after the phone boots up and finishes the setup do I get the apps from the play store.
eyeatoma said:
What would you consider a clean install? Can you import settings from your samsung profile? I generally don't transfer with the USB, but I download all my apps after the phone is booted up from the original install. Only after the phone boots up and finishes the setup do I get the apps from the play store.
Click to expand...
Click to collapse
I would start fresh, download apps manually, most have a cloud back up that they use themselves to restore once they are installed. From a settings perspective I'd do that all again fresh and let the system determine sleeping applications etc.
Basically avoid smart switch!
Tbh it's a good sort of audit process to work out which apps you have that you don't even use anymore
I don't think I was prepared for quite how big this phone is dwarfs my s10+ in size and weight