Huawei’s own developed Kirin 950 is appearing to be a very capable ‘system on a chip’. In addition to being fabricated on TSMC’s 16 nm FinFET process, the chipset will be running an octa-core processor, where one-half of the processor will comprise up of Cortex-A72 cores running at a speed of 2.4 GHz. Now, the latest benchmarks of Kirin 950 have been leaked, and from the looks of it, the chipset will outgun several upcoming high-performing SoCs.
Kirin 950 Benchmark Leaks Reveal That The Chipset Is Easily Able To Exceed The Scores Of Exynos 7420
An image originating from China reveals that Kirin 950 is not only processed on the 16 nm FinFET process, but the Cortex-A72 cores running at a clock speed of 2.4 GHz are sufficient enough to outrun Exynos 7420. Perhaps the biggest difference in performance is the presence of these four Cortex-A72, which will outclass the four Cortex-A57 cores present in Samsung’s current flagship SoC. Given below are the scores of both chipsets taken from their Geekbench benchmarking results.
- Single-core test: 1,909 points
- Multi-core test: 6,096 points
- Single-core test: 1,486 points
- Multi-core test: 4,970 points
It can clearly be seen that Kirin 950 has a comprehensive lead over Exynos 7420. However, Huawei is not going to have the last laugh just yet. Earlier, we published a report where the benchmarking results of Samsung’s Mongoose SoC were leaked, and from what we have seen, Samsung is also developing a SoC contender to take on the future mobile chipsets that are going to be found in handsets during the 12 month period of 2016.
Similar to Apple’s A8 SoC, Kirin 950 will also have an i7 co-processor present it, summarizing that Huawei is going make sufficient breathing room for the primary processors to work their magic. Since this is just another leak, we are going to have to treat it with a grain of salt and wait for more details to pour in before we can provide a more conclusive analysis regarding performance. On a side note, 2016 is going to be a fantastic year as far as mobile computing performance is concerned.