The NPU (Neural Processing Unit), as a dedicated AI acceleration chip, requires its hardware interface design to meet the demands of high bandwidth, low latency, and flexible scalability. Below are the commonly used interface signals and corresponding protocols for NPUs at the hardware level, explained with actual chip examples: 1. Memory Interface Function: Connects to … Read more
In September 2025, the AI glasses sector is heating up—Meta’s Orion 2 and Apple’s Vision Pro 2 have officially announced the integration of 12GB LPDDR5X and Micro-OLED optics, supporting on-device 7B parameter large models. Domestic companies like Starry Meizu and Rokid have also released consumer-grade AI glasses, clearly defining the three essential components: “display + … Read more
Cambricon (688256) is the first domestic AI chip stock in China, focusing on cloud training and inference chips, with extensive self-developed IP licensing and a well-established computing power layout, showing high growth potential.Innosilicon (688008) is a leader in memory interface chips, essential for AI servers relying on DDR5 interfaces, benefiting from the computing power infrastructure.Jingjia … Read more
Hello everyone, welcome to my Smart Space. Let’s embark on an immersive learning journey into embedded systems together. The Cortex-M3 is a 32-bit processor core with a 32-bit internal data path, 32-bit registers, and a 32-bit memory interface. The CM3 adopts a Harvard architecture, featuring separate instruction and data buses, allowing instruction fetching and data … Read more
NAND Flash and NOR Flash are two common types of flash memory.NOR Flash was first developed by Intel in 1988, changing the landscape that was previously dominated by EPROM and EEPROM.NAND Flash was released by Toshiba in 1989, emphasizing lower cost per bit, higher performance, and the ability to be easily upgraded through interfaces like … Read more