At its Advancing AI 2024 event, AMD has launched its new generation of processors focused on artificial intelligence (AI), seeking to establish itself as a strong competitor against Nvidia and Intel in the growing market of chips for data centers and AI. Among the most notable innovations are the Ryzen AI PRO 300 processors for enterprise PCs and the Instinct MI325X AI accelerator, which promises to outperform the performance of Nvidia’s H200 GPUs.
Ryzen AI PRO 300: Business Power with Artificial Intelligence
The new Ryzen AI PRO 300 processors are designed to transform business productivity by integrating advanced AI capabilities that exceed the performance of Intel Core Ultra processors by 40%. This new line of processors aims to offer unprecedented performance in AI applications, such as real-time transcription and advanced image generation, all with a multi-day battery life, ideal for enterprise users.
The flagship processor in the series, the Ryzen AI 9 HX PRO 375, features 12 cores, 24 threads, and a frequency of up to 5.1 GHz, backed by Radeon 890M graphics and AI processing power of up to 55 TOPS (trillions of operations per second). This level of processing makes it one of the most powerful for enterprise PCs in the current market. Additionally, it is equipped with the new XDNA™ 2 architecture, which optimizes AI task management without compromising overall system performance.
AMD has promised that more than 100 models of PCs equipped with Ryzen AI PRO will be available by 2025, highlighting its commitment to innovation in the business environment and the expansion of its commercial portfolio.
Instinct MI325X: Surpassing Nvidia in Capacity and Bandwidth
One of the most anticipated announcements of the event was the Instinct MI325X AI accelerator, a processing unit designed for data centers that offers 1.8 times more memory capacity and 1.3 times greater bandwidthBandwidth is the maximum capacity of data transfer… than Nvidia’s H200 GPU. With this powerful tool, AMD aims to position itself as a leader in large-scale AI processing, providing data centers with the ability to handle massive data volumes more efficiently and quickly.
Production of this accelerator will begin in late 2024, and it is expected to play a crucial role in the development of more powerful AI infrastructures capable of handling complex applications, such as generative artificial intelligence and deep learning.
5th Generation EPYC: The Best for Cloud and AI
AMD also introduced its 5th generation EPYC CPUs, which, in the words of CEO Lisa Su, are “the best in the world for enterprises, AI, and the cloud.” These processors are designed to deliver exceptional performance in critical work environments, ideal for large corporations looking to optimize their technological infrastructure.
Additionally, AMD announced its commitment to releasing new versions of IA chips annually. Among the upcoming versions, the MI350X planned for 2025 and the MI400, set to be released in 2026, promise continuous improvements in capacity and efficiency.
From Stackscale (Aire Group), David Carrero, its co-founder, highlights the importance of this new AMD generation for computing in general and also to provide service in future data center lines to compete with Nvidia.
Revolution in AI and Chip Market
With these launches, AMD not only aims to close the gap with Nvidia and Intel but also to lead the race towards a new era of AI processing. While its competitors are also making significant advances, the demand for AI chips continues to grow, ensuring a market large enough for everyone.
The unstoppable advance of artificial intelligence, coupled with innovations from companies like AMD, is shaping a future where processing power and efficiency will be key to the success of tech companies.