What happens after Moore’s law? Experts weigh in on the next evolution of computing

DCM Editorial Summary: This story has been independently rewritten and summarised for DCM readers to highlight key developments relevant to the region. Original reporting by The Conversation, click this post to read the original article.

For about 50 years, computing followed a steady pattern. Transistors—the building blocks of computer chips—kept getting smaller, making computers faster and more powerful. You probably used devices that improved quietly in the background, enabling better weather forecasts, realistic graphics, and the rise of machine learning. This trend became known as Moore’s Law, which observed that the number of transistors on a chip doubled roughly every two years, leading to miniaturization and consistent improvement in performance.

Today, that predictability is fading. It’s not because innovation has stopped, but because the original physical assumptions have hit their limits. Now, instead of automatic boosts in speed, the industry is turning to multiple strategies. You’ll find new materials and better transistor designs that reduce energy waste and leakage, helping maintain power efficiency. Chip layouts are changing too—components are now stacked or packed more tightly, which decreases the time and energy needed for data travel.

Another major change you’ll see is specialisation. Rather than relying on one general-purpose processor, modern systems blend CPUs for general tasks, GPUs for graphics and parallel operations, and AI accelerators for machine learning. Performance now depends on how well these components work together. All of these aim to make systems more adaptive and efficient, based on the task at hand.

You might also hear about experimental technologies like quantum and photonic processors. These aren’t meant to replace your everyday computer, but they show promise in solving tough problems such as optimisation and simulation. At events like the Supercomputing SC25 conference, there’s growing excitement about using them as co-processors—working alongside classical systems to enhance their capabilities. For regular tasks, though, improvements in traditional chips, memory, and software still bring the biggest benefits.

In this new era, computing progress is less automatic and more targeted. You’ll see bigger gains in certain applications like AI tools or scientific modelling, while general-purpose performance improves more slowly. Life after Moore’s Law isn’t about decline—it’s about adapting, making smarter hardware choices, and building software that understands those limits. The future of computing will demand more intention and smarter design, not just smaller transistors.