Advanced's Trading Volume Surges 35.49% to 53.27 Billion, Secures Ninth Spot in Market Rankings

On June 12, 2025, Advanced's trading volume reached 53.27 billion, marking a 35.49% increase from the previous day, placing it ninth in the day's stock market rankings.

AMD has introduced its latest Instinct MI355X accelerators, which the company claims outperform Nvidia's most advanced Blackwell GPUs in inference tasks. These new accelerators are designed to reduce energy and infrastructure costs, making them a competitive option in the AI hardware market.

AMD's new MI355X GPU offers significant improvements in AI performance, including double the AI FLOPs and more HBM memory compared to previous models. This enhancement allows for 40% better tokens per dollar than NVIDIA's offerings, positioning AMD as a strong contender in the AI hardware sector.

AMD's AI roadmap includes advancements in inferencing speed, which is over three times faster than the previous generation, and cost per token, which is up to 40% more efficient. These improvements are part of AMD's broader strategy to enhance its AI capabilities across GPUs, networking, and software.

In performance tests, the MI355X accelerators have shown impressive results, achieving 4.6 PetaFLOPS on their own and 36.8 PetaFLOPS in an eight-GPU platform configuration. These figures highlight the MI355X's superior performance in AI workloads, making it a compelling choice for large-scale AI applications.

AMD's new Instinct MI350 series, which includes the MI355X, is designed for large-scale AI workloads. The series features 3nm technology, 185 billion transistors, and 288 GB of HBM3E memory, providing robust support for AI tasks. The MI355X is reported to be 35 times faster than the MI300 and 2.2 times faster than Nvidia's Blackwell GPUs, further solidifying AMD's position in the AI hardware market.

AMD has accelerated the release of the MI355X to mid-2025, aiming to capture the growing demand for AI inferencing workloads. This strategic move underscores AMD's commitment to staying ahead in the competitive AI hardware landscape.