Ahead of the upcoming Vision 2024 event, Intel has released a short video on X platform showcasing a new composite chip featuring “at least ten” modules. Industry insiders have confirmed that this chip is the upcoming Gaudi 3 artificial intelligence accelerator.
The video reveals that the Gaudi 3 chip incorporates a cutting-edge design with two interconnected compute modules and eight high-bandwidth memory (HBM) stacks. This modular architecture enhances the chip’s performance while improving scalability and reliability.
Intel has disclosed that the Gaudi 3 accelerator utilizes a 5nm manufacturing process, resulting in a fourfold increase in BF16 performance compared to Gaudi 2, double the network bandwidth, and a 1.5 times improvement in HBM memory bandwidth. These advancements position Gaudi 3 as a powerful solution for handling large-scale data and complex computational tasks.
The launch of the Gaudi 3 AI accelerator represents a significant milestone in Intel's ongoing commitment to innovation in the AI sector. As artificial intelligence rapidly evolves, Intel is continuously enhancing chip performance to provide robust computing support for AI applications. Gaudi 3 will primarily support deep learning and large generative AI models, addressing the growing demand for AI computing power.
Experts in the field indicate that the introduction of Gaudi 3 will solidify Intel's standing in the AI chip market while propelling industry innovation. As AI technology gains traction, more businesses and organizations will leverage high-performance AI chips to enhance their operational capabilities and competitive edge.
Intel is expected to share more details about the Gaudi 3 accelerator and demonstrate its performance across various applications during the Vision 2024 event. The company remains committed to increasing its research and development efforts in AI, aiming to deliver advanced and efficient AI solutions.
In summary, the launch of the Intel Gaudi 3 artificial intelligence accelerator marks a significant breakthrough in the AI chip sector, offering enhanced computing power for AI applications and driving rapid advancements in AI technology. This development opens up new opportunities for innovation across the entire industry.