AMD claims its 2026 rack-scale AI solution is 4X more energy efficient than its 2024 AI platform — company says it's pacing ahead of 20X efficiency by 2030
AMD says its 2025 rack-scale AI system is 4X more energy efficient compared to its 2024 AI solution, though does not produce actual benchmark results.
AMD has announced that its 2026 rack-scale AI solution is four times more energy efficient than its 2024 platform, according to estimates released on Tuesday. The company's objective is to achieve 20 times greater efficiency by 2030, as part of its 20x2030 initiative. AMD's latest projections indicate that its rack-scale AI systems could provide the same compute power as 570 racks based on the 2024 Instinct MI300X by 2030, representing a 20X decrease in power consumption or a 20X increase in performance at the same power consumption.
The company's efficiency improvements are attributed to advancements in compute capability, memory bandwidth, and interconnect bandwidth. These enhancements result in better performance per watt, reduced wasted energy consumption, and increased performance efficiency. AMD's software optimizations also contribute to the overall efficiency gains.
It is important to note that AMD's 4X efficiency estimate should be considered an estimate rather than a direct benchmark between two commercially available rack systems. AMD's efficiency calculations are based on comparing annual representative rack configurations with a 2024 baseline using the company's performance-per-watt methodology.
Written by urgent.news from Tom's Hardware's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
