Gordon Moore co-founder of Intel discovered that since the invention of integrated circuits the number of transistors had doubled every year. Moore produced an article in the Electronic magazine titled ‘Cramming More Components Onto Integrated Circuits’ explaining his findings. His observation became widely accepted in the industry and came to be known as Moore’s Law.
The short term cramming of components was expected to continue if not increase. Yet the long term rate of increase was a little uncertain but was to remain almost constant. Originally, Moore predicted that the number of transistors in an IC would double every year. In 1975 Gordon Moore’s prediction was revised at the International Electron Devices Meeting. It was determined that after the year 1980 it would slow down to doubling every two years.
The extrapolation of this data has been used in the semiconductor industry for many years to direct long term planning and set targets for research and advancement. From your laptop, your camera and your phone – any digital electronic device is heavily linked to Moore’s Law. Moore’s Law became somewhat of a goal for the industry to reach, ensuring timely progression in technology. Society has benefited greatly from this advancement in all areas, such as education, health, 3D printing, drones, government etc.
At the 1975 IEEE International Electron Devices Meeting, Moore outlined several factors he believed were contributing to this exponential growth:
- As techniques improved, the potential for defects has dramatically decreased.
- This combined with an exponential increase in die sizes meant that chip manufacturers could work with larger areas without losing reduction yields
- Development of smallest dimensions achievable
- Conserving space on a circuit known as circuit cleverness – optimizing how clever components are arranged and eventually find the optimum use of space
Major Enabling Factors
Moore’s Law wouldn’t be viable without a few innovations by scientists and engineers over the years. This is the timeline of the factors that enabled Moore’s Law:
|1947||John BardeenWalter Brattain||Built first working transistor|
|1958||Jack Kilby||Texas Instruments||Patented the principle of integration and created the first prototype of an integrated circuit and commercialized them|
|Kurt Lehovec||Sprague Electric Company||Invented a way to isolate components on a semiconductor|
|Robert Noyce||Fairchild Semiconductor||Created a way to connect components on an IC by aluminum metallization|
|Jean Hoerni||Planar technology based the improved version of insulation|
|1960||Group of Jay Last’s||Fairchild Semiconductor||Made the first operational semiconductor integrated circuit|
|1963||Frank Wanlass||Frank Wanlass|
Invented complementary metal-oxide-semiconductor (CMOS)
|Allowed extremely dense and high-performance IC’s|
|1967||Robert Dennard||IBM||Created dynamic random-access memory (DRAM)||Enabled the possibility of fabricating single transistor memory cells (led to the invention of flash memory by Fujio Masuoka from in the ’80s allowing low-cost high capacity memory in many devices)|
|1980||Hiroshi ItoC Grant Wilson J. M. J. Frechet||Invented chemically-amplified photoresist (5-10x more sensitive to UV light) – IBM introduced to DRAM productions mid-1980’s|
|IMB||Created deep UV excimer laser photolithography||Enabled the smallest components of an IC to shrink even smaller (1990 800nanometer – 2016 10 nanometers)|
|Late 1990’s||Innovations of interconnects from chemical-mechanical polishing or chemical-mechanical planarization (CMP)||Enables improved wafer yield by additional layers of metal wires, closer spacing and lower electrical resistance (not a direct factor in smaller transistors, but a major development for improved IC’s)|
Is Moore’s Law Still True
It’s commonly asked and debated whether Moore’s Law is still true. While there is disagreement amongst experts as to the answer to this – it is commonly agreed that it is no longer the driving force in the transistor industry.
In the past, expansion of storage and computation capabilities was based on aggressive feature scaling, manifested in Moore’s Law. However, scaling will not be able to address the upcoming needs in IC performance and utilization of energy resources. Not only that, but advancements have decreased and other technology options to keep Moore’s Law alive are being researched.
Since 1998, the industry has produced roadmaps for semiconductors using Moore’s Law to drive advancements. In 2016 the final roadmap was produced. The industry is no longer centered around Moore’s Law but it is outlined by a strategy that could be called ‘more than Moore’s Law’. It is based around research and development of the needs and applications of chips, rather than scaling sizes. The application of chips varies from smartphone and laptops to artificial intelligence and data centers.
When it comes to the future development of technology, Moore’s Law was great as it allowed everyone in the industry to have a common heartbeat, work together and create a bit of healthy competition between companies. Not just that, but consumers and other developers knew what to expect in advancements.
As Moore’s Law is finishing up it gives the industry a chance to explore new avenues and get creative. Looking at the physical architecture, such as getting rid of designs from the 1940s could unlock the potential for higher efficiency. By having to redesign the basic architecture of computing, programmers will have to alter their old habits and adopt a new way of thinking –
As the industry moves on to ‘more than Moore’s Law’ strategy, the question is will companies be able to progress at the same rate of growth and scale? Many believe that the rate of advancement won’t be the same, due to the simple fact that companies will have to work together in a new and complicated way, without the common heartbeat which kept all the research and development plans in sync. Therefore, advancements that benefit all could become less common.
Moore’s Second Law
The primary driving force of economic growth is the growth of productivity. Moore’s Second Law (also known as Rock’s Law) looks at the economic flip side of semiconductor production.
The prediction was made in the 1960s by Arthur Rock – a businessman and early investor in tech companies such as Intel. It simply states that the cost of semiconductor fabrication also mimics exponential growth – it doubles every four years while the cost of a product for consumers was halved. The price of a fabrication plant for semiconductors had already reached around 14 Billion US Dollars in 2015.
Moore’s Second Law evaluates the ongoing growth of financial investment required in the semiconductor industry. As advancements improved it was possible for manufacturers to have the ability to create better machines to automate the production line. The automation process has created lower priced products for the consumer as the hardware created has lower labor costs. Newer and popular products being sold means more profit to invest in developing new innovative designs and devices of even higher capabilities.
The cost of manufacturing a single unit is always decreasing while money being invested is constantly increasing to continue research and development. At some point, Moore’s First Law (the number of components on an IC) and Moore’s Second Law (the expenses of producing IC’s) will collide – as the rising costs of manufacturing will reach a plateau and become too expensive to maintain and profit from.