High Snapdragon 820 single-core benchmark leak bodes well for the Xiaomi Mi5

High Snapdragon 820 single-core benchmark leak bodes well for the Xiaomi Mi5
So far, the highest Snapdragon 820 single-core score we've seen from the alleged Xiaomi Mi5, or project Gemini, was 2162, which is still a far cry from, say, Apple's breathtaking single-core achievements with the A9 in the iPhone 6s, which can hover around the 2500 mark. 

Yesterday, however, a leakster account known for braking some god insights already, tipped a Snapdragon 820 single-core tally of 2300 points, which, while still not on the level of Apple A9, is still a very high score, and bodes well for the performance of phones with the new Snapdragon inside. One of these phones will most likely be the Xiaomi Mi5, and this 2300 score may have been achieved precisely with it, given that the rumor originates from the Weibo account of the leakster, and the fact that Mi5 tips had popped up from there before. 

In any case, next year's battle for the heart of Android flagships is shaping up to be between Exynos 8890 and Snapdragon 820 - given that the Exynos just scored a record on AnTuTu, while the 820 may be killing it Geekbench, we can't wait for the first phone(s) with them to arrive come January, as the tipster implies.

source: Leaksfly (Weibo) & Geekbench

FEATURED VIDEO

29 Comments

1. hafini_27

Posts: 944; Member since: Oct 31, 2013

Good enough.

2. zeeBomb

Posts: 2318; Member since: Aug 14, 2014

Pretty damn impressive single core, considering it has 4 cores too.

3. Finalflash

Posts: 4062; Member since: Jul 23, 2013

No its a 2 big 2 small setup so likely not going to be that great. Seeing as how Android scales off more cores really well, it won't be winning again the Exynos too easily. Once again though, CPU matters little compared to the GPU which is what people should really be focusing on.

4. w1000i

Posts: 235; Member since: Jul 22, 2015

Wrong, IT IS 4 identical COREs , with clock speed difference. And when the difference is only clock speed, you can gain back the performance if you can clock it high or save battery by clocking it lower

5. hung2900

Posts: 966; Member since: Mar 02, 2012

It's still unknown whether it's the same or not. The "little" core can be another revision to be benefited from the big.LITTLE implementation. Like Cortex A53, the 2.2 GHz cores in high-end MTK Helio X10 SOCs is different from the 1.3 GHz core in low-end SOCs (and something like S410).

9. j2001m

Posts: 3060; Member since: Apr 28, 2014

Very true, the single core of the iPhone A9 is crap it needed to be more like 3000 and even more so when you look at the the A9x single core, the A9 is very low and crap for 2 cores for the new amd custom design CPU, I be thinking this is due to clock speed Please note that the new iPhone speed is linked to the new cheap ver of ufs HD they are using that can r and w at the same time and as nothing to do with the A9, so if your phone uses Emmc next year you get nothing from the new CPU if it's sd 820 or any of the other CPUs as you have a big bottleneck to the cpu

8. Guaire

Posts: 850; Member since: Oct 15, 2014

Probably 2 cores clocked lower to avoiding overheating. So overclocking them isn't a good idea.

12. alex3run

Posts: 715; Member since: May 18, 2014

The cache setup is different. That leads to the difference in performance per clock for big and little clusters.

24. may_czos

Posts: 954; Member since: Nov 22, 2014

I wish they used 3 fast cores and one slow core (which is still plenty for less demanding tasks) with good clock management and the ability to stop unused cores individually.

25. ibend

Posts: 6747; Member since: Sep 30, 2014

it isnt identical if clock speed is different.. lower clock speed will be small core.. but who care, all 4 core will active when they run heavy task

6. Hexa-core

Posts: 2131; Member since: Aug 11, 2015

great package for a low price! woah

7. Hexa-core

Posts: 2131; Member since: Aug 11, 2015

great package for a low price! woah

10. jove39

Posts: 2137; Member since: Oct 18, 2011

Wow...that is impressive score for kryo...not sure why author said "still not on the level of Apple A9"...Did you meant despite of a9's 6 wide twister core compared to kryo (which is 3 wide), single core score difference is of only 200 points?

27. ibend

Posts: 6747; Member since: Sep 30, 2014

nobody care about single core since no app run on single core, except fruit lover, since apple always have high single core score because they use huge core

11. jove39

Posts: 2137; Member since: Oct 18, 2011

Read this section of article about geekbench benchmark. http://www.pcworld.com/article/3006268/tablets/tested-why-the-ipad-pro-really-isnt-as-fast-a-laptop.html?page=3 Its good article, if you like to read it all.

14. alex3run

Posts: 715; Member since: May 18, 2014

The conclusion about A9X speed came from the vacuum. No real tests, just words. The only test (editing 4K movies) failed and nobody knows exactly why.

15. jove39

Posts: 2137; Member since: Oct 18, 2011

I suggest you read entire article. Test include tabletmark, 3dmark and file compression comparison as well. And one test based on MS Excel.

17. alex3run

Posts: 715; Member since: May 18, 2014

Sorry, maybe I wasn't too attentive... Thank you for noticing =)

18. alex3run

Posts: 715; Member since: May 18, 2014

Well, I've reread the article and... did not find the tests.

20. alex3run

Posts: 715; Member since: May 18, 2014

So thank you once gain. I've didn't notice that the article has several pages.

13. alex3run

Posts: 715; Member since: May 18, 2014

Still behind Exynos and even Mediatek with Huawei in more important multicore.

23. may_czos

Posts: 954; Member since: Nov 22, 2014

Single core (and GPU) score is much more important.

26. ibend

Posts: 6747; Member since: Sep 30, 2014

why??

28. alex3run

Posts: 715; Member since: May 18, 2014

Apple told you that? =) In real life in our Universe (not Apples one) multi core performance does matter more than single core performance. But yea, the GPU is the most important part.

29. Busyboy

Posts: 731; Member since: Jan 07, 2015

Nah, most apps use only one core. Only demanding apps use more.

16. vuyonc

Posts: 1088; Member since: Feb 24, 2014

Still no GPU specific benchmarks yet for the SD820 or E8890 :/

22. TheGunnyPT

Posts: 252; Member since: Feb 12, 2015

Yep, Really worried about this.. If we are getting 4k phones this year, I really want to know which GPU will power them..

21. fyahking

Posts: 1146; Member since: Jan 28, 2015

Apple A9 is not quite there in multicore, yet.

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.