Arm Unveils 'Armv9 Edge AI Platform' Running On-Device AI

British semiconductor design company Arm unveiled a new central processing unit (CPU), the 'Arm Cortex-A320,' capable of running artificial intelligence (AI) models with over one billion parameters in on-device environments, along with the 'Armv9 Edge AI Platform' equipped with this CPU, on the 27th.

Hwang Seon-wook, President of Arm Korea, explained the background of the product launch at a press conference held at The Plaza Hotel in Jung-gu, Seoul, stating, "As AI models become more complex, they demand higher performance and more power. Flexibility to continuously update AI devices has become necessary, and the importance of edge AI is increasingly emphasized."

Edge AI is an infrastructure that combines mobile communication networks and AI computing. Compared to large-scale AI data centers, it is known to be advantageous in terms of low latency, enhanced security, and privacy protection. It is especially gaining attention in the Internet of Things (IoT) sector, such as autonomous vehicles navigating factory sites and smart cameras that adjust functions through software updates.

The new CPU, Arm Cortex-A320, operates as the "heart" of the Armv9 Edge AI Platform. The product improves machine learning performance by 10 times compared to its predecessor, the Cortex-A35. Arm explained that it provides advanced security features, making it useful for edge devices operating in exposed environments or handling sensitive data. This product is scheduled for mass production next year.

The Armv9 Edge AI Platform equipped with the Cortex-A320 can run large language models (LLM) and small language models (SLM) for agent AI. By performing decision-making closer to the data collection point, it can also reduce latency.

Arm also announced that it is expanding the computing library set for developers, 'Arm Kleidi,' into the IoT field.

This content was produced with the assistance of AI translation services.