Samsung's next flagship processor features an integrated NPU for AI

Samsung Electronics is the latest mobile systems-on-chip provider to start using dedicated hardware for device-based machine learning. The Exynos 9 Series 9820, which will undoubtedly be used in the international versions of the Galaxy S10, features a Neural Processing Engine (NPU) designed to deliver AI performance up to seven times faster than the S10 9810. NPUs can be used to speed up tasks such as image processing and AR.

Huawei was the first chip designer to announce a SoC with an NPU when it unveiled the Kirin 970 last year, and Apple quickly followed up with the Aion Bionic; Both companies have focused more on the performance of AI with their new chips Kirin 980 and A12 Bionic. Qualcomm, whose Snapdragon 845 chip is in the US version of the Galaxy S9 and virtually all of the other flagship Android products other than Huawei's around the world, does not yet manufacture AI hardware, although it does not include any hardware. he cited the D45's combination of DSP, processor and graphics processor, offering three times more AI performance than the previous 835.

In addition, Samsung claims that the Exynos 9820 fourth-generation custom processor cores will provide a 20% increase in conventional single-core performance and a 15% increase in multicore or 40% energy efficiency. The graphics processor uses the new ARM Mali G76 cores, already discovered on the Kirin 980, for 40% better performance or 35% higher energy efficiency.

According to Samsung, the Exynos 9820 will be ready for mass production by the end of the year, which means that the supposed launch of the Galaxy S10 will take place at the end of February. The question now is whether Qualcomm will follow the rest of the industry by focusing on AI for its successor 845.