Nvidia AI GPU may soon enter MediaTek’s flagship mobile SoCs. This move will be an important step for Nvidia, as it demonstrates that established desktop GPU architectures can be adapted to smartphone SoCs. Nvidia’s new strategy aims to increase its AI computing power and help the company recover from its financial problems in 2022. Integrating Nvidia’s GPU into MediaTek’s SoC will benefit from improved AI performance, currently lagging behind Qualcomm’s Snapdragon 8 Gen 2.
PC users have already been able to experiment with the AI processing power provided by integrated tensor cores in GeForce RTX GPUs over the past few years, but DigiTimes reports that smartphone users may soon be able to benefit from similar technologies. Industry groups indicated. Nvidia’s AI GPU could be integrated into MediaTek’s 2024 flagship mobile SoC.
This is a great opportunity to prove that established desktop GPU architectures can be adapted to smartphone SoCs. AMD and Samsung attempted a similar feat with the Exynos 2200 SoC connecting the RDNA 2 GPU, but the performance paled in comparison to Qualcomm’s Snapdragon or Apple’s Bionic implementations. Whether Nvidia can accommodate an architecture as complex as Ada Lovelace and miniaturize it to provide any meaningful performance advantages over competing solutions remains to be seen.
Apart from the GPU side of things, the MediaTek SoC also benefits from enhanced AI performance. Currently, MediaTek’s AI implementation in the flagship Dimensity 9200 lags behind what Qualcomm offers with the Snapdragon 8 Gen 2. If MediaTek can gain even a tiny bit of performance compared to Qualcomm’s flagship solutions, more smartphone brands will find it more attractive. Use Dimensity SoCs, helps MediaTek conquer the premium arena.
DigiTimes noted that MediaTek denied such rumors, but they seem rather strange considering the fact that MediaTek has already paired Nvidia AI GPU with some Arm-based SoCs designed for Chromebooks.