Forget about the Note 10 and get excited for the Galaxy S11 with the first Snapdragon 865 benchmark

Forget about the Note 10 and get excited for the Galaxy S11 with the first Snapdragon 865 benchmark
If you've been following the news as far as high-end smartphone chips are concerned, you might be a little disappointed to see both the Galaxy Note 10 and Note 10+ powered by the "old" Snapdragon 855 rather than the hot new 855+ SoC. Of course, "international" variants of both new handsets are packing Exynos 9825 processors, but that's a different discussion.

What potential US buyers should keep in mind is that the Snapdragon 855+ silicon brings modest speed improvements to the table, as suggested by Qualcomm itself and confirmed in early benchmarks of the gaming-centric Asus ROG Phone 2 and Xiaomi Black Shark 2 Pro. Naturally, that will absolutely not be the case when the Snapdragon 865 is released, and we already have some benchmark results of what we presume to be a reference device of sorts to prove it.

Who's ready to get excited about the Galaxy S11?


Yes, it's early days, and apart from some vague camera upgrade ambitions and a pompous codename, we really don't know anything of substance about Samsung's "next big thing"... after the as-yet-unreleased Note 10 and Note 10+.

 

But we can safely assume that the Galaxy S11 will be using Qualcomm's next flagship SoC in the US. According to the typically reliable @UniverseIce over on Twitter, the Snapdragon 865 is powering a mysterious "Qualcomm Kona" device that's been able to rack up as much as 4160 single-core and close to 13000 multi-core performance points in recent Geekbench tests.

As the leaker highlights, these scores might not be representative of the 865's "final production version", but they should still be taken seriously, already hinting at some massive gains over both the Snapdragon 855 and 855+. We're talking a single-core bump of around 700 points compared to the best official results posted by the 855 on commercial devices, as well as close to 2500 points gained in multi-core capabilities. That would roughly equate to a 20 percent improvement in speed, although the number could easily jump to 30 percent or even more between now and the launch of the Galaxy S11 and other spring 2020 Android flagships.


While this Qualcomm Kona reference device apparently runs Android 10 on the software side of things, we're pretty sure the OS is not fully optimized yet. It's also important to point out the Snapdragon 865 SoC was capable of producing the aforementioned results in combination with a modest 6 gigs of RAM. So, yeah, those numbers are certainly not final and they don't accurately reflect what this beast will be able to do out in the real world.

A heavyweight contender for Apple's best chipset?


The answer to that question seems to be yes, but only for the time being. This early, unrefined version of the Snapdragon 865 is indeed faster (at least on paper) than the A12 Bionic found inside the iPhone XS and XS Max, but we'll have to wait and see what the A13 powering the next iPhone generation can pull out of its hat. If recent history is any indication, Apple shouldn't rest on its laurels just yet.


After all, the A12 Bionic only improved the multi-core total of the A11 Bionic inside the iPhone X, 8, and 8 Plus by around 10 percent. It looks like the Cupertino-based tech giant will need a much more significant jump this time around to stay ahead of the competition, although as far as single-core performance is concerned, its domination is probably safe.

In case you're wondering, both the Apple A13 and Qualcomm Snapdragon 865 are expected to use something called extreme ultraviolet (EUV) lithography to enhance the capabilities of the 7nm process on which the A12 Bionic and Snapdragon 855 are currently built. But while Apple is pretty much guaranteed to partner with TSMC yet again, Qualcomm might rely on Samsung's expertise to manufacture its next big thing for use on 2020's best Android handsets. 

FEATURED VIDEO

48 Comments

1. ph00ny

Posts: 2026; Member since: May 26, 2011

Honestly i'm more excited about potential Exynos with AMD GPUs. Hopefully it's the Navi GPUs

2. Back_from_beyond

Posts: 1377; Member since: Sep 04, 2015

That's unfortunately gonna take at least another 2 years.

8. AbhiD

Posts: 720; Member since: Apr 06, 2012

Qualcomm too uses AMD tech in it's GPUs. So Samsung won't have any specific advantage. Moreover it's at least 2 - 3 years away at best.

31. oldskool50

Posts: 1323; Member since: Mar 29, 2019

Yes they will. There is a difference in Qualcomm using a piece of architecture from AMD vs AMD actually talkign a Radeon chipset and placing it on the die with the Exynos. What QC did with AMD is the same thing they do as a ARM licensee. They use the architecture in some form or fashion. But what Samsung will be doing is far more.

36. AbhiD

Posts: 720; Member since: Apr 06, 2012

You must be super naive to think that Samsung will be offering Radeon with smaller die... LMFAO! Do you even realise how difficult it is to make full fledged desktop GPUs usable for smartphone usage? NVIDIA tried too hard and long and still failed royally. Let's discuss it after 2-3 years where this project ends up

46. oldskool50

Posts: 1323; Member since: Mar 29, 2019

What? The Adreno is a near desktop class GPU. You're saying you cant have a low level awesome GPU? You can have a nice dedicated GPU that doesnt require active cooling. You do know that right?

38. ph00ny

Posts: 2026; Member since: May 26, 2011

AMD hasn't touched adreno in ages. You must be the naive one to think that mobile application will be configured similar to desktop variants or that we're talking about full on mobile configuration. If it's based on RDNA/Navi, it probably will reduce the number of cu accordingly

41. vincelongman

Posts: 5677; Member since: Feb 10, 2013

Exactly, AMD sold ATI's mobile GPU division literally over 10 years

6. Sam1ne

Posts: 38; Member since: Apr 07, 2019

Um..There's way more to benchmarks than just the geekbench multicore..also, the single core is arguably more important and there the Exynos 9250 and A13 will reign supreme. I have a feeling the biggest strength of the SD 865 could be the GPU

9. AbhiD

Posts: 720; Member since: Apr 06, 2012

To burst your bubble, Samsung's Exynos has been failing in speedtests for 2 years now despite having higher single core scores. Because it's single core numbers never come to life in real life, just limited to benchmarks. Go and study basics on Anandtech. On the other hand Snapdragon has been reigning supreme from last 2 - 3 years.

10. eggimage

Posts: 83; Member since: Jan 19, 2012

This is so not true with mobile even for today. Single core performance still weighs heavily in the mobile realm. Also, snapdragon has not been leading, but still trailing behind apple’s chips, it’s only recently when they have finally caught up, but with tradeoffs on battery life, while apple’s chips still lead far ahead overall if you normalize the power consumption

20. AbhiD

Posts: 720; Member since: Apr 06, 2012

Did you even care to read what i wrote? I clearly said that single core scores of Exynos mean NOTHING! As equivalent Snapdragon has been beating it at speedtests from long. It's obvious that Exynos single core score doesn't translate to real life performance. As for Apple, wake me up the day when their A series processors start sucking lesser battery and stop throttling

33. oldskool50

Posts: 1323; Member since: Mar 29, 2019

He specifically talked about the Samsung Exynos speeds and benchmarks vs the Snapdragon. I see you one of those people who have reading comprehension issues. He never even mention the A-series.

37. AbhiD

Posts: 720; Member since: Apr 06, 2012

If he didn't even mention about A series, then what is this kiddo ? "Also, snapdragon has not been leading, but still trailing behind apple’s chips, it’s only recently when they have finally caught up, but with tradeoffs on battery life, while apple’s chips still lead far ahead overall if you normalize the power consumption"

47. oldskool50

Posts: 1323; Member since: Mar 29, 2019

I was responding to him talking to you. I told egg, YOU never mentioned ed the A-series.

44. sgodsell

Posts: 7207; Member since: Mar 16, 2013

First of all benchmarks don't mean jack in the real world. Apple's Ax suck in real world. They over heat and throttle down, especially if you stress them. But not the Snapdragons. Plus I would love to see how any iPhone fairs when it's running two tasks simultaneously. But alas no iPhone can do that. You would think split screen multitasking in 2019 would be possible on any device. Let us know when Apples iPhones can run more than one thing at a time. Also eggimage multicore is the real metric, not single core.

24. Sam1ne

Posts: 38; Member since: Apr 07, 2019

The Exynos 9810(especially) and 9820 were problematic because if throttling. But the 9250 sounds like they finally might got it right and actually release the full potential of the M4 CPU.

32. oldskool50

Posts: 1323; Member since: Mar 29, 2019

FALSE. The single core is pointless, when the operating system and the application are all multi-threaded. You have it in total reverse. That is why on mobile SoC vs desktop processors, they do things different On a mobile SoC you have multiple cores that run at different maximum speeds. In a dual quad design, 4 cores could be lower clocked for low end tasks, while 4 cores are cloecked higher for more throughput. Apple does the same with 6 cores 2 may be slower 4 will be faster as an example. Desktop cores all run at basically the same clock, but any of the 4 or more can be dedicated to a task at anyone time. Then when you add hyper-threading, you get more. So how the hell you think a single core score is beter than multi-core on platforms that use multi-thread OS, multi-thread apps, and multi-thread SoC/CPu? You are completely clueless in what you claim The only time single core scores mattered is when, all chips were simply a single one core schip. That si no longer the case. All chips are multi-core, because Application are multi-threaded. To say otherwise is complete ignorance. The only reason you trying to hold on to single score is because the A-series still benches higher. Complete asinine.

40. c.m.s

Posts: 235; Member since: Dec 10, 2017

Single core is pointless? I guess you do not work as a developer. Single thread performance is certainly not pointless. Far from all tasks can be split up into subtasks that you can process in parallell, specially the kind of workload you have an a mobile plattform. You always have certain problem you do need to process sequentially and then it does not matter how many cores/threads you have, the single core performance is what matters here, so single core is definitely still important.

42. vincelongman

Posts: 5677; Member since: Feb 10, 2013

Geekbench single core is pointless Remember that the 9820 crushed the 845 in Geekbench single core, yet it struggled to beat the 835 in other benchmarks and real world performance A proper single core benchmark like SPEC2006 shows that Arm are ahead of Samsung in single core

48. oldskool50

Posts: 1323; Member since: Mar 29, 2019

I do. Still doesn't change any fact. Apps that need more power are coded from the beginning to be multi threaded. Even apps that dont need the extra power could be still threaded over multiple cores, because it's the OS that controls what happens after the app launches. Just because an app is not specifically coded to take advantage stage of extra cores, doesn't mean it won't. What it dows mean, is it won't take full advantage.

7. AnTuTu

Posts: 1602; Member since: Oct 14, 2012

First you spend months covering rumors and what not for Note 10. Then one day later after the release "Forget about Note 10". What are you trying to do man?

11. eggimage

Posts: 83; Member since: Jan 19, 2012

You’re comparing SD865–a chip that’s not even in the market yet—with apple’s chip from 2018, and criticizing apple for only making “small improvements”?? Also A12 didn’t just gain 10% in performance, its neural engine is literally 900% faster than A11. Also you completely disregarded power consumption. One major reason A12’s cpu performance seemingly gained little improvement is because it put more focus on battery life, also its Tempest power efficient cores are much faster than before too. A12’s performance was already over a year ahead of SD855’s, let alone the A12X inside iPad Pro. Now next year’s SD865 will only able to rival A12 from 2018. I don’t see how that’s even a competition.

14. Back_from_beyond

Posts: 1377; Member since: Sep 04, 2015

That neural engine is useless for most common uses of the phone. The CPU didn't improve much and they were right to point that out. And the A12X is a tablet only SoC so there's no valid comparison to an SD865 that's meant for smartphones. The 855 keeps up with the A12 just fine, as no doubt the 865 will do with the A13.

43. vincelongman

Posts: 5677; Member since: Feb 10, 2013

A12X? That a tablet SoC, why mention that lol The 855 already beats the A12 in efficiency

12. iloveapps

Posts: 855; Member since: Mar 21, 2019

So sad with Note 10 having a year old soc and mid range specs compared to other android. And there’s no way android chips can beat apple’s A series chips. Though android had more ram, they are still behind with performance, battery and optimization.

15. Back_from_beyond

Posts: 1377; Member since: Sep 04, 2015

And yet they do keep up with the A12 and in some areas even do better.

19. pimpin83z

Posts: 469; Member since: Feb 08, 2019

LMAO! To quote Ratchet from Transformers: Revenge of the Fallen, "This fool is terribly misinformed."

21. wickedwilly

Posts: 667; Member since: Sep 19, 2018

Behind on performance and battery…that is completely the opposite of what I get and by far the majority of tests I have seen.

22. sissy246

Posts: 7070; Member since: Mar 04, 2015

LMFAO And here you are, trying so hard to put down what you wish mommy would buy you. Go watch some videos and you will see that the s10+ keeps up or within a couple seconds and in some cases bets the iPhone in opening a app. Now days they are very close to each other so BFD if one is a couple seconds faster. You act like it is minutes difference between the two.

* Some comments have been hidden, because they don't meet the discussions rules.

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at https://www.parsintl.com/phonearena or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit https://www.parsintl.com/ for samples and additional information.