14 DAY TRIAL // Qualcomm’s Snapdragon 820 is one of the most anticipated high-end silicon platforms expected to make a debut towards the end of the year. And in order to build hype around the platform, the US chip maker has been teasing some of the most important aspects related to the SoC.
The Snapdragon 820 is the first to come from the Qualcomm garden that will take advantage of the company’s own custom cores dubbed Kyro. The silicon is of a quad-core variety with a 2.2GHz clock and has been built using 14nm FinFET technology.
Anyway, Qualcomm has been seemingly testing several versions of the Snapdragon 820, and now, according to Tech Grapple, it has already reached version 3.
A few months ago, the first version scored only 714 in single-core performance and 1942 in the multi-core one, but these low results can be attributed to the fact that the Kyro cores were running at less than half of their maximum 3GHz capacity (at 1.46GHz to be more exact).
The Snapdragon 820 gets benchmarked again
Then came the second version, which scored a lot better, with 1732 points in single-core and a score of 4970 points in multi-core tests. Well, this week we get news of the third version of the Snapdragon 820 being out and about, which, according to the current report, scored an impressive 2032 points in single-core and 5910 in multi-core.
Compared to the Exynos 7420 inside the Samsung Galaxy Note5, which got 1431 points in single-core and 4717 points in multi-core, the chip fared pretty well.
However, in contrast with Apple’s new A9 chipset fueling the iPhone 6s, the Snapdragon 820 only managed to beat it in multi-core (A9 scored 4375) but failed to eclipse it in single-core, where the Apple silicon got 2529 points.
A few days ago, information related to Samsung’s next-big silicon piece, the Exynos 8890, broke out showing us that, at 2.4GHz clock speed, the chipset could score 2304 points in single-core and a massive 8938 points in muti-thread.
So it does appear at this point that Qualcomm is facing some solid competition, if we’re to take into account these synthetic benchmark results. But more often than not, they aren’t so telling. Real-life performance is.
