How far can we push computing?
MIT Technology Review Explains: Let our writers untangle the complex, messy world of science and technology to help you understand what’s coming next. You can read more from the series here. On the evening of December 29, 1959, at an American Physical Society meeting at Caltech, the physicist Richard Feynman gave a lecture called “There’s…
In December 1959, physicist Richard Feynman delivered a groundbreaking lecture at an American Physical Society meeting in California, positing that the laws of physics permitted the manipulation and control of matter at an atomic scale. He famously argued that it would be possible to store an entire Encyclopaedia Britannica on the head of a pin or pack the entirety of all written books within the space of a single dust speck.
Feynman's comments highlighted the vast potential for innovation in the field of computing, and indeed, for the next several decades, advancements in technology continued to push the boundaries of what was possible.
Transistors, the fundamental building blocks of computer chips, underwent a series of miniaturizations, with their size continually halving. This led to a doubling in the density of chips, enabling a computer that once filled an entire room to now comfortably fit within a watch. However, this steady progress has since plateaued, and the clear indication of this change lies in the massive investments made by leading data center operators.
Over the past year, these companies have spent more than $700 billion on computing infrastructure, a substantial increase from the $350 billion they had spent just five years prior. This significant expenditure represents a reliance on brute force—adding more of what is already known to be effective—to boost performance.
Despite the stagnation in the evolution of computer hardware, the field of computing has entered a new era dubbed the "intelligence age." The once-dominant trends that fueled the remarkable improvements of the 20th-century information age are now showing signs of decline. Moore's Law, which describes the exponential increase in transistor density on silicon, has slowed down; Dennard's Law, which maintains consistent power consumption as transistors become smaller, has ended; Koomey's Law, tracking the energy efficiency of computation, is beginning to falter; Kryder's Law, which concerns the packing density of magnetic storage, has slowed to a crawl; and Keck's Law, which outlines the exponential rise in the amount of information transmitted through conventional optical fibers, is starting to flatten.
These scaling laws, while not representing fundamental physical laws, reflect the culmination of human-engineered systems and provide a clear picture of the room left for further innovation.
Written by urgent.news from MIT Technology Review's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
