Kirin 990 Benchmarks & Performance Review

In September 2019, Huawei launched its new flagship devices Mate 30 & 30 Pro. These smartphones are powered by Huawei’s latest Kirin 990 chip.

With Kirin 990, Huawei has gone with a Dual-SOC approach by making one chip with 5G Modem for 5G smartphones and the other one for non-5G devices.

Apart from the 5G Modem, these chips differ in some other aspects as well. The 5G variant of Kirin 990 is fabricated using TSMC’s 7nm+ (N7+) EUV process while the 4G variant is manufactured using the older 7nm (N7) process that was also used in the previous-gen Kirin 980.

Kirin 990 5G also has faster clock speeds, thanks to the enhancements in the new N7+ Process and it also has a better AI Accelerator. It is a massive chip with a die size of 113.31 mm². For comparison, Apple A13 Bionic and Snapdragon 855+ have a size of 98.48 mm² and 73.27 mm² respectively. The 4G variant of Kirin 990 is much smaller with a size of 90 mm².

Before we go in depth in the review and benchmarks, here is a quick look at the pros. and cons. of this SOC.

  • Extremely Powerful
  • Excellent for Gaming
  • Extremely Power Efficient
  • Great AI Performance
  • Just like the last year’s Kirin 980, Huawei once again made use of the same LPDDR4x Memory, Cortex-A76 CPU, and the Mali-G76 GPU. But this time around, the CPU clocks are higher and the Graphics Processing Unit features more cores.

    With these previous-gen architectures, Huawei won’t be able to compete with Samsung’s Exynos 990 and Qualcomm’s Snapdragon 865. Both these SOCs feature LPDDR5 RAM. Samsung’s custom CPU Core and the Cortex-A77 Core in Snapdragon 865 outperform Kirin 990 in the CPU department. The story is similar in the GPU department where Qualcomm’s Adreno 650 and the Mali-G77 GPU in Exynos 990 are way ahead of Kirin 990’s Mali-G76 GPU.

    However, Kirin 990 also has a small advantage over Snapdragon 865 and Apple’s A13 Bionic as it uses TSMC’s N7+ Process which is slightly better than the N7P Process used in the other two SOCs. If you want to learn more about the fabrication nodes, you should go through this article .