Moore’s law in computing

Moore’s Law is one of the most influential observations in the history of computing and technology. Formulated in the mid-twentieth century, it describes the exponential growth of computing power and the miniaturisation of electronic components, particularly transistors, over time. The law has served as both a descriptive observation and a guiding principle for the semiconductor industry, driving innovation, performance improvement, and cost reduction in electronic devices for more than five decades.

Origin and Formulation

Moore’s Law was first proposed by Gordon E. Moore, co-founder of Intel Corporation, in an article published in Electronics magazine on 19 April 1965. Observing the trend of integrated circuit development, Moore noted that the number of transistors on a microchip had roughly doubled every year since the invention of the integrated circuit in 1958, and he predicted that this trend would continue for at least the next decade. In 1975, Moore refined his forecast, stating that the number of transistors on a chip would double approximately every two years.
This exponential growth implied that computing performance would increase dramatically while costs per transistor would decrease correspondingly. Moore’s observation was not a physical law but an empirical projection based on technological progress, manufacturing techniques, and industry innovation.

Technological Implications

The implications of Moore’s Law have been profound. The continual doubling of transistors has enabled the design of smaller, faster, and more energy-efficient microprocessors and memory devices. Over time, this has transformed computing from large, costly mainframes into compact, affordable personal computers and later into mobile devices, wearables, and embedded systems.
Key developments driven by Moore’s Law include:

• Increased processing speed: Greater transistor density allows faster switching and higher computational throughput.
• Reduced cost per computation: As chips become more powerful and efficient to manufacture, computing becomes cheaper and more accessible.
• Miniaturisation: The ability to place billions of transistors on a single chip has enabled the miniaturisation of electronic devices.
• Expansion of computing applications: From scientific research to entertainment and artificial intelligence, powerful yet affordable processors have revolutionised multiple fields.

Semiconductor Scaling and Manufacturing Advances

The fulfilment of Moore’s Law has relied heavily on continuous advancements in semiconductor fabrication technology. The principal mechanism enabling transistor miniaturisation has been scaling, the process of reducing the size of semiconductor components while maintaining or improving their performance.
Over the decades, successive technological milestones have been achieved:

• Micrometre to nanometre scaling: Transistor sizes have shrunk from tens of micrometres in the 1960s to single-digit nanometres in the 2020s.
• Planar to 3D structures: The introduction of FinFET (Fin Field-Effect Transistor) and Gate-All-Around (GAA) transistors improved performance and power efficiency.
• Advances in lithography: Optical lithography techniques, such as Extreme Ultraviolet (EUV) lithography, have allowed precise patterning at atomic scales.
• Material innovations: Beyond traditional silicon, materials like silicon-germanium, gallium nitride, and carbon nanotubes are being explored to sustain miniaturisation.

These innovations have collectively allowed manufacturers to maintain the pace predicted by Moore’s Law for several decades, even as the physical and economic challenges of further miniaturisation have intensified.

Economic and Industrial Impact

Moore’s Law became a self-fulfilling prophecy in the semiconductor industry. Companies such as Intel, AMD, NVIDIA, TSMC, and Samsung set their research and development roadmaps in accordance with Moore’s prediction, striving to maintain the expected doubling rate of transistor density.
This relentless pursuit of performance improvement established the rhythm of “technology nodes”, typically every 18–24 months, which signalled the introduction of new manufacturing processes. The effects were visible across the global economy:

• Consumer electronics: Each new generation of processors enabled faster, cheaper, and more capable devices.
• Scientific and industrial computing: Supercomputers, data centres, and AI systems benefited from exponential performance gains.
• Economic growth: The reduction in computing costs spurred innovation in software, telecommunications, finance, and healthcare.
• Energy and efficiency gains: Newer chips often provided higher performance per watt, supporting the growth of mobile and cloud computing.

Limitations and Physical Challenges

As transistor sizes approached the atomic scale, maintaining the rate of progress predicted by Moore’s Law became increasingly difficult. Several physical and economic constraints have emerged:

• Quantum effects: At nanometre scales, electrons can “tunnel” through thin barriers, causing leakage currents and reliability issues.
• Thermal dissipation: Greater transistor density increases heat generation, posing challenges for cooling and efficiency.
• Fabrication costs: The cost of building semiconductor fabrication plants (“fabs”) has risen sharply, reaching tens of billions of dollars.
• Diminishing returns: The performance gains from scaling alone have slowed, necessitating new approaches such as parallelism and heterogeneous computing.

By the mid-2010s, industry leaders, including Intel, acknowledged that Moore’s Law was slowing down, with the doubling period extending beyond two years. Despite this, innovation continued in other dimensions of computing performance.

Beyond Moore’s Law: New Paradigms

The perceived slowing of Moore’s Law has not signalled the end of technological progress but rather a transition to new computing paradigms. Researchers and engineers have explored alternative strategies to sustain performance growth:

• Parallel and multi-core architectures: Instead of increasing clock speed, processors incorporate multiple cores to perform simultaneous computations.
• Specialised accelerators: Graphics Processing Units (GPUs), Tensor Processing Units (TPUs), and AI accelerators enhance specific workloads.
• Quantum computing: Explores computation using quantum bits (qubits), offering exponential performance in specific problem domains.
• Neuromorphic and bio-inspired computing: Mimics the structure of the human brain for efficient pattern recognition and learning.
• 3D chip stacking and advanced packaging: Increases data transfer speeds and reduces energy consumption by vertically integrating chip layers.

These developments represent an evolution rather than a replacement of Moore’s Law, extending its spirit of continual improvement through new technologies and architectures.