Applying Moore's Law: Revolutionizing Computing Power And Beyond

Moore's Law is a prediction made by Intel co-founder Gordon Moore in 1965 that the number of transistors per silicon chip would double every year. Moore's Law has been applied in the semiconductor industry to guide long-term planning and set targets for research and development. The law has been applied to other areas of technology too, such as hard disk drive areal density and network capacity. However, it is not a law of physics but rather an empirical relationship. Moore's Law has been questioned in recent years, with some believing it has already ended, and chipmakers are finding it increasingly challenging to continue to apply the law.

Moore's Law and data growth

Moore's Law, named after Intel co-founder Gordon Moore, is a computing term that originated in the 1970s. The law states that the number of transistors on an integrated circuit (IC) doubles about every two years, leading to ever-more powerful computers at lower prices. This has fuelled technological progress and economic growth over the last 50 years, with advancements in digital electronics such as microprocessor prices, memory capacity, sensors, and digital cameras strongly linked to Moore's Law.

The exponential growth in data has outperformed Moore's Law, and the gap has become more apparent in recent years. This "tsunami of data" has led to a shift towards optical cores in backbone communication networks to support the increasing demand for data transmission and processing. Optical interconnect networks offer lower latency, higher throughput, and reduced power consumption, making them essential in handling the exponential growth of data.

The challenges of data growth have also impacted data centre architectures, requiring new approaches to storage, computation, and communication. Distributed networks, custom processing elements, and synchronised operations on massive datasets are now necessary to manage the explosion in data volumes.

While Moore's Law is slowing, it is being supplemented by strategies that exploit systemic complexity. Custom devices, multi-die designs, 3D memory stacks, and synchronised software stacks are some of the approaches being explored to continue advancing computational performance.

Moore's Law and hardware

Moore's Law states that the number of components on a single chip doubles every two years at minimal cost. It is not an actual law of physics but an observation and projection of a historical trend. It is an experience-curve law, a type of law that quantifies efficiency gains from experience in production.

The law was formulated by Gordon Moore, the co-founder of Fairchild Semiconductor and Intel, in 1965. He observed that the number of components per integrated circuit had doubled every year between 1960 and 1965, and projected that this rate of growth would continue for the next decade. In 1975, he revised his forecast, predicting that the number of components would continue to double but at a rate of once every two years.

Moore's Law has been used in the semiconductor industry to guide long-term planning and set targets for research and development. It has also been a driving force of technological and social change, productivity, and economic growth.

The primary method of Moore's observation has been to make wires and transistors that transmit and process information smaller and smaller. This has resulted in an explosion in computing power. However, this explosion in the size of computing components cannot continue indefinitely due to the laws of physics. As components get smaller and more numerous, they generate more heat, and eventually, the chip will melt.

The physical limits of Moore's Law are expected to be reached in the 2020s. Chip-makers are facing increasing costs to continue meeting the industry standard, as well as difficulties in cooling an increasing number of components in a small space.

Despite these challenges, Moore's Law has remained fairly accurate for nearly 50 years. In 2024, chip-makers can put 50 billion transistors on a chip the size of a fingernail.

Moore's Law and the future of technology

Moore's Law, an empirical relationship, has been the driving force behind the semiconductor industry's development over the last 50 years. It states that the number of transistors in an integrated circuit (IC) doubles about every two years.

The future of Moore's Law

The future of Moore's Law is uncertain, with some experts declaring it dead and others believing it will continue. While it has been predicted that Moore's Law will end by around 2025, technological progress is expected to continue in other areas, such as new chip architectures, quantum computing, and AI and machine learning.

The impact of Moore's Law ending

The end of Moore's Law will mark a shift from an era of brute-force solutions to one where creativity and innovation will be paramount. It will also impact semiconductor design, as designers will need to develop products with added value derived from design rather than manufacturing.

Strategies for continued performance improvements

To maintain performance improvements, a multi-pronged approach is necessary, including architectural specialization, advanced packaging technologies, and the development of CMOS-based devices that extend into three dimensions. Additionally, there is a focus on improving materials and transistors to enhance performance and create more efficient underlying logic devices.

The role of new models of computation

Quantum and brain-inspired computing technologies have gained attention due to their rapid improvement; however, they are not replacements for digital electronics. These technologies expand computing capabilities in specific areas, such as pattern recognition and solving complex problems, but digital computing remains essential for accurate and reproducible calculations.

Architectural specialization

Architectural specialization, such as the use of accelerators, is a near-term strategy to continue performance improvements. This approach tailors hardware to specific applications, optimizing hardware and software for particular tasks to enhance performance and energy efficiency.

Data movement challenges

Another challenge for future digital technologies is the cost of data movement, which already dominates electrical losses. As transistor sizes decrease, the energy efficiency of wires does not improve, resulting in higher energy costs for data movement than for computational operations.

Photonics and rack disaggregation

The integration of photonic technologies, which offer higher bandwidth density and energy efficiency, can address data movement challenges. Photonic interconnect technologies can provide system-wide disaggregation, enabling flexible sharing of hardware resources in data centers.

CMOS replacement

The development of new devices, such as improved transistors or digital logic technology, can significantly reduce energy consumption. However, this requires fundamental breakthroughs in materials and a deep co-design approach that considers manufacturing challenges to ensure economic manufacturability at scale.

Advanced manufacturing

To meet societal needs and expectations, new devices and computing paradigms must provide exponential improvement while being economically manufacturable. This may require a technological shift on par with the transition from vacuum tubes to semiconductors, demanding a strategic foundation for change.

Moore's Law and the semiconductor industry

Moore's Law, an empirical law of economics, has been applied in the semiconductor industry to guide long-term planning and set targets for research and development. It has also been used to create a roadmap for the industry's future, with companies striving to increase processing power.

The law, named after Intel co-founder Gordon Moore, states that the number of transistors on a microchip doubles about every two years with minimal cost increase. In 1965, Moore observed that the number of transistors on an integrated circuit had been doubling every year and projected this rate of growth to continue for at least another decade. He revised this prediction in 1975, stating that the number would double every two years.

Moore's Law has been a driving force in the semiconductor industry, influencing technological and social change, as well as economic growth. It has led to advancements in digital electronics, such as the reduction in microprocessor prices, the increase in memory capacity, the improvement of sensors, and the number and size of pixels in digital cameras.

However, there are challenges to sustaining Moore's Law. The physical limits of the law are expected to be reached in the 2020s as it becomes harder to make transistors smaller. Additionally, the cost of research and development, manufacturing, and testing has been increasing with each new generation of chips. Despite these challenges, Moore's Law has had a significant impact on the semiconductor industry and continues to guide its future development.

Moore's Law and the end of physical limitations

Moore's Law is the observation that the number of transistors in an integrated circuit (IC) doubles about every two years. It is not a law of physics but an empirical relationship. It is an experience-curve law, a type of law that quantifies efficiency gains from experience in production.

The law was formulated by Gordon Moore, the co-founder of Fairchild Semiconductor and Intel, in 1965. Moore initially projected that the number of components per integrated circuit would double every year for the next decade. In 1975, he revised his forecast to doubling every two years.

The prediction has been used in the semiconductor industry to guide long-term planning and set targets for research and development. It has also been a driving force of technological and social change, productivity, and economic growth.

However, there are physical limitations to Moore's Law. The primary technical challenge is the design of gates as device dimensions shrink. Controlling the current flow in the thin channel becomes more difficult as transistors get smaller. This leads to issues such as source-to-drain leakage and limited gate metals and channel materials.

Other approaches that do not rely on physical scaling are being investigated, such as spin-based logic and memory options, tunnel junctions, and advanced confinement of channel materials via nano-wire geometry.

The end of Moore's Law will have significant implications. It has been a driving force of technological progress, and its end will impact the development of new technologies. Additionally, the cost of research and development to uphold Moore's Law has been rising, and the fabrication plants for creating advanced chips are becoming prohibitively expensive.

The search for successors to silicon chips will require significant research and development. While there are potential alternatives, such as quantum computing and carbon nanotube transistors, there is no obvious replacement for the promise of Moore's Law.

Frequently asked questions