Smartphone Makers are Increasing RAM to Accommodate On-Device AI

AI-driven RAM boost in smartphones: Apple’s iPhone 16 lineup and Google’s Pixel 9 series are leading a trend of increased RAM in smartphones, primarily to support on-device artificial intelligence features.

• Apple’s iPhone 16 models now come with 8GB of RAM, up from 6GB in last year’s base models, as discovered by MacRumors.
• Google’s Pixel 9 series has also seen a RAM increase, with 12GB now being the minimum across all models.
• The RAM boost is largely attributed to the memory-intensive nature of AI models, which require quick access to stay responsive.

The relationship between AI and RAM: On-device AI features demand significant memory resources to function efficiently and provide a seamless user experience.

• AI models need to be constantly loaded in RAM for quick response times, as loading from long-term storage would result in noticeable delays.
• Even small AI models, like Microsoft’s Phi-3 Mini, can occupy 1.8GB of space, necessitating more RAM to avoid impacting other smartphone functions.
• Google’s Android VP, Seang Chau, explained that the Pixel 8 Pro’s additional 4GB of RAM allowed it to better handle the Gemini Nano AI model compared to the standard Pixel 8.

Google’s approach to AI and RAM: The tech giant has been transparent about the challenges and solutions related to implementing AI features on their Pixel devices.

• Initially, Google didn’t enable local AI features on the standard Pixel 8 due to RAM limitations.
• The company later allowed Gemini Nano on the Pixel 8, but only for users willing to run their phones in Developer Mode, acknowledging potential impacts on user experience.
• With the Pixel 9 series, Google increased RAM across the board to accommodate AI features without compromising overall phone performance.

Implications for user experience: The RAM increase is expected to benefit both AI capabilities and general smartphone performance.

• Google’s Stephanie Scott stated that users can expect better performance and improved AI experiences from the additional RAM in Pixel 9 phones.
• The extra RAM is also intended to future-proof devices for upcoming AI advancements, although larger AI models may consume more memory over time.
• Apple hasn’t explicitly stated RAM requirements for its Apple Intelligence features, but all compatible devices have at least 8GB of RAM.

Parallel trends in the laptop market: The push for more RAM to support AI features is also evident in the personal computer industry.

• Microsoft has set a minimum requirement of 16GB of RAM for laptops to be considered “Copilot Plus PCs,” capable of running local Windows AI features.
• Apple is rumored to be planning RAM increases in its next generation of laptops, moving away from the long-standing 8GB default.
• Hugging Face CTO Julien Chaumond suggests that 6-8GB of RAM is the “sweet spot” for keeping an AI model always loaded alongside other OS functions.

Apple’s approach to AI and hardware: While not explicitly discussing RAM, Apple has hinted at the hardware requirements for its AI features.

• Apple AI boss John Giannandrea mentioned that factors like “bandwidth in the device” and neural engine size affect the feasibility of AI features on different iPhone models.
• Craig Federighi, Apple’s VP of software engineering, acknowledged that RAM is one of the components in the “total” package required for AI functionality.

Future implications: The trend of increasing RAM in smartphones is likely to continue as AI becomes more prevalent in mobile devices.

• The relatively small 2GB RAM increase in the iPhone 16 lineup may be just the beginning of a larger shift towards more memory-intensive devices.
• As AI models evolve and grow in complexity, future smartphones may require even more RAM to maintain performance and support advanced AI features.
• This trend could lead to a new era of mobile computing where AI capabilities become a primary driver of hardware specifications and upgrades.