Artificial intelligence (AI) is gradually becoming a part of every aspect of our lives, and even our most commonly used and inseparable devices are incorporating it. Mobile phones are expected to increasingly rely on AI for various tasks, both common and innovative; however, this development will come at a cost. An Omdia analyst has made an interesting prediction for the near future: AI in smartphones will significantly increase the need for storage capacity, reaching 500 GB or more by 2025.
Currently, Google’s Pixel 8 Pro is the most advanced AI phone on the market. Although its features are continually improving, and new ones are added slowly, the device’s Tensor G3 chip only handles some functions with its neural processing unit (NPU), while others occur online. This is expected to change by 2024, with even more drastic changes expected in 2025, resulting in greater hardware resource requirements.
Mobile phones are entering a new era with AI, and significantly more storage capacity will be necessary. Omdia’s principal researcher, Yeon Seung-hoon, has raised concerns in his latest report, stating:
“Next year, with the integration of artificial intelligence features in smartphones, there will be growth opportunities for storage. To drive AI operations, including learning and inference, more than 15% of total smartphone storage must be utilized.”
Considering the storage capacities of current smartphone models vary between 128 GB and 1 TB with NAND Flash — with lower capacities being more popular — the loss of at least 15% of that capacity, along with the increasing size of apps and operating systems, presents a concerning situation.
By 2025, at least 500 GB or more of storage will be required. Both Google’s push in the AI sector and Samsung’s upcoming announcements on January 17 and Apple’s in late 2024 will offer mixed cloud and system-on-chip (SoC) services. However, by 2025, it is predicted that SoCs will have much more powerful NPUs, introducing heavy algorithms on devices to ensure better user privacy. As a result, companies like Apple, Google, Samsung, and possibly Huawei, are expected to adopt local generative AI, increasing the demand for storage capacity.
Early reports suggest that the minimum size to handle AI will be 512 GB, with 1 TB being ideal. This means powerful AI phones with 2 TB and even 4 TB storage capacities will likely emerge. Samsung is already planning to introduce UFS 4.0 optimized for AI in January, targeting this increased storage demand. Therefore, either we’ll need to transition to a new generation of NAND Flash technology or the price of smartphones with local AI capabilities will rise significantly, further segmenting the market.
The integration of AI in smartphones will lead to increased storage needs for proper functionality, with a minimum of 500 GB required by 2025.