Anyway, let’s check out how the Snapdragon 810 in the One M9 performs.
There’s a couple of things to note here before I dig into the numbers in the benchmarks above. Firstly, results seemed to fluctuate more than normal while running our set of browser benchmarks in Chrome. I’m not sure why this is, but it didn’t seem to be related to temperature throttling or the SoC itself, as the Snapdragon 810-powered G Flex 2 I also have on hand didn’t exhibit the issue. Due to the fluctuating results, the in-browser benchmark numbers for the One M9 are an average of ten runs. Secondly, I actually achieved greater system and CPU performance on average benchmarking the G Flex 2, despite both phones having the same SoC with the same clock speeds. I suspect this is due each company using a different performance governor in the OS, but it’s something I’ll explore further in my imminent review of the G Flex 2. Anyway, the One M9’s implementation of the Snapdragon 810 either outperforms or matches the Samsung Exynos 5433 we saw in the Galaxy Note 4. As a refresher, the Exynos 5433 features the very same big.LITTLE octa-core as the Snapdragon 810, but clocked slightly lower on both core clusters. It’s not surprising to see the One M9 outperforming the Exynos Note 4 by a small margin, around 7% on average, which is slightly lower than the average difference in CPU clock speeds (15%). As the Exynos 5433 outperforms the Snapdragon 805 in CPU-heavy tasks by 7%, it’s not surprising to see the Snapdragon 810 also outperform the Snapdragon 805, in all but Basemark OS II 2.0. The average margin comparing these two Qualcomm SoCs is 12% in favour of the Snapdragon 810. Compared to the One M8, the One M9 is 26% faster in CPU-bound tasks, which is a respectable gain on silicon released only a year ago.
