Intel introduces AI to manage power systems, Faster and more energy efficient

According to PC World, at this year’s Hot Chips, Intel presented for the first time its AI-powered energy management system employed in the “Meteor Lake” CPU. This innovative system harnesses artificial intelligence to discern a user’s behavioral patterns, seamlessly transitioning between high and low power states at optimal intervals.

When computers grapple with intensive tasks, to achieve peak performance, they often allocate augmented power to the processor. Consequently, the processor operates at elevated frequencies to expedite task completion. Concurrently, it’s imperative for the CPU to accurately predict the completion timeline of these tasks, facilitating a swift transition to a power-conserving mode, a process known as Dynamic Voltage and Frequency Scaling (DVFS). While Intel has utilized a myriad of techniques to regulate active and idle power states in its preceding CPU generations, the imminent Meteor Lake CPU heralds the advent of AI technology to refine power management.

As delineated by Intel, the power management system of the Meteor Lake processor can adeptly “comprehend” and prognosticate user actions, like the sequence of opening, scanning, and subsequently closing a webpage before continuing their browsing expedition. This algorithm is equivalently adept at other tasks. Noteworthy, however, is its self-evolving nature. The behavioral patterns it extrapolates are far more nuanced and precise than those previously hardcoded by Intel.

Leveraging the Meteor Lake CPU, artificial intelligence will dictate the power allocation for the CPU under specific workloads, transmuting power into frequency to unlock superior performance. Intel asserts that this AI prowess can amplify system responsiveness by 35%, assisting the CPU in determining the apt moment to transition from a high-power to a low-power state, thereby curtailing power consumption by an impressive 15%.