On-device artificial intelligence (AI) capabilities are backed by current-generation chips such as the Snapdragon 8 Gen 3, with future System on Chips (SoCs) expected to enhance these capabilities further. While on-device AI imposes significant storage and computational requirements, there is room for improvement in feature development. Nonetheless, a study strongly argues against Samsung and Apple incorporating extensive computational dependencies in the Galaxy S24 and iPhone 16, citing the current limitations in AI content generation and the escalating semiconductor costs.
To bolster on-device AI functionality, upcoming iPhones and Android smartphones will necessitate substantial storage and RAM capacities.
As smartphone chipset production transitions to advanced processes like TSMC’s 3nm ‘N3E, prices are projected to escalate. Speculations suggest that the Snapdragon 8 Gen 3 could surpass the cost of the Android 8 Gen 2, with Qualcomm’s priciest SoC slated for release in the subsequent year. Mizuho Securities anticipates a lenient approach from phone manufacturers like Samsung and Apple amidst these price hikes, potentially resulting in the Galaxy S24 and iPhone 16 lagging in upgraded memory provisions.
The limited availability of compelling on-device AI content implies that manufacturers might lack the incentive to equip the Galaxy S24 and iPhone 16 with extensive RAM, especially given the rising semiconductor costs highlighted in Revegnus’ post on X referencing Mizuho Securities’ forecast. Samsung purportedly announced that the upcoming Galaxy S24 lineup will only support 12GB RAM, exclusively in Samsung’s GSM Plus and SSM Ultra configurations.
According to Mizuho Securities, enhancing storage capacity is imperative for on-device AI, yet laptop manufacturers are expected to adopt a conservative approach due to the projected surge in memory silicon prices and the scarcity of engaging on-device AI content.
Apple is anticipated to maintain 8GB RAM for the “Pro” models of the iPhone 16 while reportedly introducing significant microphone enhancements to enhance speech input for Siri and other Artificial Intelligence-Generated Content (AIGC). Apart from the substantial RAM requirements, on-device AI will necessitate a minimum of 15% of the available memory to support myriad operations.
The potential integration of this technology into the Galaxy S24 series by Samsung remains uncertain, although rumors suggest the development of new UFS 4.0 flash storage tailored for application execution. In conclusion, there are conflicting reports regarding the trajectory of smartphones with AI capabilities, and as the new season unfolds, more insights into this evolution may emerge.
