Apple flash: Soon, our smart gadgets will be more intelligent like Own AI.

Apple flash: Soon, our smart gadgets will be more intelligent

Animated By Arabian Space Science

Apple's Breakthrough in Memory Technology for AI on Smart Devices

Apple has made a significant breakthrough in addressing a major obstacle in the integration of artificial intelligence (AI) on portable devices. The challenge lies in the limited memory capacity of smartphones, which hinders the implementation of large language models and other AI capabilities.

In a recent publication on the arXiv pre-print server, Apple researchers unveiled their innovative solution. By leveraging data transfers between flash memory and DRAM, they have devised a method that enables smart devices to effectively run powerful AI systems.

The key advantage of this approach is its ability to accommodate AI programs that are twice the size of a device's DRAM capacity. Additionally, it significantly enhances the performance of both CPU and GPU operations. CPU operations can experience a boost of up to 500%, while GPU processes can be accelerated up to 25 times compared to current methods.

The researchers achieved these remarkable results by developing an inference cost model that aligns with the behavior of flash memory. This model guides the optimization process in two crucial aspects: minimizing the amount of data transferred from flash memory and reading data in larger, more contiguous chunks.

This breakthrough holds immense potential for expanding the capabilities of our smart devices. Imagine engaging in detailed and natural language conversations to explore academic or personal topics. Furthermore, the integration of global health databases could enable real-time monitoring of vital signs, alerting users to potential health issues. Additionally, comprehensive translation services between multiple languages could become seamless, facilitating effective communication across borders.

Apple's utilization of large language models and natural language processing has already revolutionized our interactions with technology. With this latest advancement in memory technology, we can anticipate even greater progress in bridging the gap between AI and portable devices.