🕒 Loading time...
🌡️ Loading weather...

Ai Mainstream

AMD CEO welcomes us to the “YottaScale “era”—Lisa Su says AI will need YottaFLOPS of compute power soon

At the start of CES 2026, Dr. Lisa Su, the CEO of AMD, opened the event with a keynote address. During her speech, Su highlighted the increasing demand for massive computing power in the realm of artificial intelligence (AI), referring to it as the “Yottascale” era. This new era is estimated to require computing power approximately 10,000 times greater than what was needed for previous generations of AI technology.

Dr. Lisa Su emphasized that the AI industry is on the brink of a significant transformation that will necessitate a substantial increase in computational capabilities. She pointed out that by the end of the decade, the world is projected to require up to 10 YottaFLOPS (a one followed by 24 zeros) of computing power. This figure represents a substantial leap from the global AI compute capacity observed in 2022, which was approximately one zettaflop (a one followed by 21 zeros).

Acknowledging the current scarcity of available computing resources relative to the demand for AI applications, Su outlined AMD’s forthcoming strategy to tackle this challenge. She expressed astonishment at the unprecedented scale of computational requirements posed by the evolving landscape of AI technology.

To address this monumental demand for computing power, AMD’s strategy will involve integrating various components such as CPUs, GPUs, networking infrastructure, and software systems. These integrated solutions are designed to enhance the scalability and efficiency of AI infrastructure significantly.

Dr. Lisa Su underscored the critical importance of AI as a transformative technology with far-reaching implications across diverse industries. She stressed that AMD is fully committed to advancing AI capabilities and described it as their top priority moving forward.

During her keynote presentation, Dr. Lisa Su introduced several new products from AMD aimed at revolutionizing AI technology. These products included cutting-edge offerings such as the MI455 GPU, EPYC Venice CPUs, and Helios AI-rack scale solutions, all designed to deliver remarkable advancements in performance and energy efficiency.