Free «Limits to Moore's Law» Essay Sample

Introduction

The semiconductor industry credits its success from the decisive contributions that were provided by Moore. Moore observed that the number of transistors that were present in an integrated circuit doubled after two years. Doubling of transistors has brought about crucial developments in the increase in processing speed, the memory capacity of storage devices and the number of pixels in cameras. As Moore predicted, in 1970 the processing speed of computers ranged from 740 KHz to 8MHz. Between the year 2000 and 2009, the processing speed ranged from 1.3 GHz to 2.8 GHz (Lazakidou, 2009). However, scientists argue that Moore’s law will reach its limits in the next ten years to come. According to Michio Kaku, finding the replacement of silicon chip will cause the collapse of Moore’s law since when processors reach 5 nm, they will overheat thus reducing their overall efficiency. This is the reason why companies manufacturing semiconductors decided to start manufacturing multi-core chips and 3D transistors. In addition to this, as the transistors become smaller the cost of producing transistors will increase. The limitations of Moore’s law have led to the development of new technologies that aim at increasing the efficiency of silicon transistors. However, these technologies have faced competition from molecular, quantum and optic computing, and this has contributed to the improvement of business processes and scope of information technology.

Description of the Technology

Moore’s Law argues that the processing speed and the overall processing power will double after every two years. In addition to this, this law states that the number of an affordable CPU’s would double after every two years. Gordon Moore, the founder of Intel Company, observed that the number of transistors of a 1-inch diameter of silicon doubled after every x number of months. Scientists argue that the number of months that are applicable to this law change according to changes in the market of microprocessors. Considering changes in the market of microprocessors, some scientists say that it takes 18 months for the number of transistors or processors to double while other scientists say that it takes 24 months. Moore wrote an article “Cramming More Components onto Integrated Circuit” and stated that, as circuits became more complex, the cost of producing the circuit increased (Huff, 2008). He added that when techniques of producing components of the circuit increased, costs of creating complex circuits would go down. Moore also predicted that ten years from 1965, the optimal number of components in integrated circuits would shoot from 50 to 65,000. This prediction was true since, from 1975, the number of components in integrated circuits had risen to about 65,000 components.

Moore’s Law is about to reach its limits. This is because scientists argue that the development of single atom transistors and processors will most probably end Moore’s Law. The single atom processor will have a single phosphorus atom. These processors will have a width of 0.1 nanometers, and it would be, therefore, hard to reduce the size of these processors as Moore predicted. The single atom transistor must be kept cold by using liquid nitrogen since it heats at a fast rate. Another limit to Moore’s Law is that as the transistors continue getting smaller, the cost of producing these transistors will become more expensive (Null & Labor, 2010). This will cause manufactures to incur more expenses, and thus they will not prefer to produce the small chips. When the size of the semiconductors shrinks from 20nm to about 18 nm nodes in 2014, the semiconductor manufacturing tools will become too expensive, and the volume of sales and production will not cover them. Companies such as Advanced Micro Devices went bankrupt after they failed to cover the costs of producing small chips. Theorists also argue that exponential growth in every industry eventually ends. They give an example on how engineers have failed to increase the speed of aircrafts and railroads further, and thus there is a high chance that Moore’s Law will reach its limit.

Along the process, as manufactures of transistors continue producing smaller transistors and processors, the laws of physics will eventually intervene. Electron tunneling will occur, and thus electrons will be moving off the atom of the transistor, and this will reduce the overall efficiency of the transistor. In addition to this, small transistors emit a lot of heat; and thus, when many transistors are installed on a silicon chip, the silicon chip will reduce its overall efficiency (Sexana, 2009). According to research conducted by IBM, the trend of the past 30 years of microelectronics has been to increase their speed and density by reducing the size of their components. IBM argued that this trend would most probably end since energy barrier will result. Scientists are considering an option of abandoning silicon and semiconductors while manufacturing personal computers. This is because the processing speed of personal computers has been increasing at a reducing rate. Instead, scientists have proposed to move from the old silicon semiconductors and produce carbon-based graphene or transition metal oxide. This will end Moore’s Law since manufactures will no longer produce silicon transistors.

Potential Applications

The limit of Moore’s Law has caused the development of several potential applications. Manufactures have started developing 3 D transistors in an attempt of reducing the high power consumptions caused by the small transistors and high heat that they emit. Previously, manufacturers were using extreme ultraviolet technology to produce 32 nanometers silicon chip. However, these chips turned out to be more expensive than the previous 90 nm chips produced; due to this, most firms producing chips have opted to use 3D-through-silicon technology to produce chips. By means of using this technology, firms could produce chips that consume 20% of power less than the previous chips (Hubner & Beckner, 2010). Manufacturers have taken advantage of less power consumption and piled different components in the chip, and this has improved its overall efficiency. Intel took advantage of 3D transistors and developed Ivy Bridge chips. The new tri-gate transistors that Intel introduced have 3 dimensional silicon fins that rise vertically from its silicon substrate. A key feature of the 3 D transistors is that they have a gate wrapped around their fins. The current in the transistors is controlled using the gate, and this gate increases the amount of current flowing in the transistor when it is on and reduces the amount of current flowing from the transistor when the transistor is off.

The Ivy Bridge is 22-nanometers in width, and this will enable the transistor to move to smaller geometrics and ensure processors use less power and become faster. The 3D transistors also allow the chips to operate at lower voltage and have low power leakages (Zelkowitz, 2011). The transistors will also have an advantage of faster switching speed. However, Intel claims that the production of Ivy Bridge will most probably increase its production costs by 2 to 3% since it will need to acquire new equipment that it will use for the manufacture of these transistors.

The Limitation of Moore’s Law has also led to the development of nanowire transistors. These transistors have made possible for engineers to produce reprogrammable circuits. The nanowires are made of germanium core and silicon shells. Nanowires are crossed with lines of metal electrodes, so as to form a grid. An advantage of these transistors is that they have the capability of maintaining the states “on” and “off” without considering whether the power is supplied or not (Hubner & Beckner, 2010). Therefore, they have the capacity of developing low-power sensors for efficient collection of data. Research has proved that these circuits conserve ten times more power compared to the circuits made by traditional silicon material. In addition to this, nanowires prevent the leakages of any electric current; therefore, improving the overall efficiency of processors.

Limitations of Moore’s law also drove computer engineers to implement parallel computing. In this form of computing, processing cores can work together in parallel, so as to solve computing instructions simultaneously. This has enabled the use of non-local resources while doing computing functions. Computer users can use resources in a wide area network when their resources are scarce, when they use parallel computing (Hubner & Beckner, 2010). Parallel computing allows for multi threading. However, a major drawback of parallel computing is that its programs are difficult to write. This is because the concurrency of parallel programs introduces potential software bugs. Most companies have been making multi core processors that run in parallel, since they discovered that serial processors produced a lot of heat and consumed a lot of power. Software engineers claim that programmers, who write programs that run on parallel computing, should put into consideration paradigm that concerns the structure of the hardware in these computers. They should put into consideration how processors and computer memories are connected to each other.

Technologies that will Compete with the Potential Applications of the Limitations of Moore’s Law

The potential applications that resulted from the limitations of Moore’s Law will receive stiff competition from quantum computers, molecular computers and optical computers. A quantum computer uses quantum mechanical phenomena to represent data. Quantum computers encode information as quantum quabits or quantum bits and the information can exist in the form of superposition. Quabits exist in the form of ions, atoms, electrons or photons, and they work on control devices, so to act as computer processor and memory (Williams, 2011). Quantum computers are able to use parallelism while performing mathematical computations and thus they are able to work on million computations at once. These computers also use the concept of entanglement, and this helps them to process data efficiently. This concept argues that when an outside force is applied on an atom, the two atoms can become entangled, and the second atom may assume the properties of the first atom. By means of using this principle, scientists are able to know the value of a quabit. Researchers claim quantinum computers will replace silicon chips.

Quatinum computers will offer stiff competition to silicon transistors computers since they will be able to carry out complex mathematical functions. In addition to this, they will consume less power and carry out computations at a faster rate than the normal silicon computers. Although silicon chips have introduced 3D transistors that have enabled fast processing of computer information, these transistors still represent data in the form of bits. In contrast, quantum computers represent data in the form of both quabits and bits and this will help in ensuring fast transmission of data (Williams, 2011). This will improve the overall efficiency of quantum computers compared to the silicon computers. Quantum computers will also have higher computing power than the normal silicon computers; thus, even if manufactures of silicon computers improve the speed of processors, they will never match the computing power of quantum computers.

Molecular computers will also offer stiff competition to the silicon computers despite the recent developments in the transistors technology. Researchers from Michigan Technological University and Japan have succeeded in manufacturing a molecular computer. This computer can mimic the human brain, repair itself and replicate the inner mechanisms of the normal human brain (Feringa & Browne, 2011). It uses organic molecules instead of silicon. Scientists built these computers by placing molecule made of oxygen, nitrogen, chlorine and carbon on a gold substrate. Hence, it has the ability of switching among the four conducting states, and this feature makes molecular computers more efficient compared to the binary silicon computers. It is remarkable that molecular computers have the ability of self-organizing its molecular layer and healing themselves just like the way human brains regenerate themselves. Molecular computers are small and allow for easy and wide networking; due to this, they will offer stiff competition to the normal silicon computers. A significant advantage of molecular computers is that they are able to use the versatility and parallelism of the human brain to perform complex mathematical computations. This is because their organic molecules facilitate their electrical impulses to travel through dynamic and neutral paths that communicate collectively and constantly with each other. Molecular computers will offer stiff competition to silicon computers due to several reasons. These computers have the ability of solving problems whose algorithms people do not know. They will have the capability of finding the cure of cancer since scientists have been able to use them to simulate the complexity of cancerous cells. These computers are also faster that silicon computers as they use four conductors to transmit data while silicon transistors only use two bits.

Optical computers also threaten to offer stiff competition to the developments in silicon computers. These computers use photons in visible light beams to perform digital computations instead of utilizing electronic current. The speed of light is ten times more than the speed of electricity; therefore, optic computers will be ten times faster than the normal electronic computers. An advantage of visible light beams is that they can pass through each other without interfering with their own paths (Zelekowitz, 2011). One successful application of optical engineering is the development of digital communications in the manufacture of fiber optic cables, which are used to transmit electronic impulses in the form of light beams. Intel used optic computing to develop Light Peak technology. Light Peak enables personal computers to connect with other external peripherals using high-speed optical cables. These cables are able to transmit data 20 times faster than the normal USB cables. This proves that optic computers will give the normal silicon stiff competition if they are introduced into the market. When manufacturers start developing optical silicon processors, they will successfully reduce the cost of manufacturing personal computers, further to reducing their overall size. Optical computing will also enable companies that have large servers to save on costs that they would incur, when purchasing bulky cables that transmit their processed data. Bulky cables are hard to coo, since they require constant cooling mechanism to ensure that they operate efficiently. Optic cables that contain waveguides, which will ensure that voluminous data is transmitted by using thin optic fibers, can replace these bulky cables. IBM recently decided to use optical interconnects in their servers since they discovered that these servers ensured faster transmission of data. Optic computing will offer stiff competition to silicon computing since it will enable the transmission of data at a faster speed and less consumption of electric power.

Effect of the Competition on Businesses

The technological competition between silicon computers, optical, molecular and quantum computers will have several effects on businesses. It will increase the speed at which businesses carry out transactions. Since producers of silicon transistors used in silicon chips know that they will face stiff computers from the emerging technology introduced by quantum computing, they have tried to produce processors that have a high speed. Intel successfully introduced Ivy Bridge that promises to bring a revolution in 3D transistors. Processors from Ivy Bridge are fast, and thus this will increase the speed in which business do transactions such as updating inventory information, simulating the shortage in production that a firm may face and producing payroll sheets. In addition to this, companies have been able to produce timely financial reports since high competition in the technology of manufacturing transistors has encouraged innovation in the development of programs that can operate in parallel (Huff, 2008). High competition has also benefited businesses since it has encouraged the development of innovative computers that can solve many organizational problems. The molecular computers will have the capacity of thinking like human beings, and thus they will be able to solve critical organizational problems within a short time. The competition in the development of small transistors has significantly reduced the size of computers.

Most companies are considering the introduction of new technologies that will overcome the limitations of Moore’s Law and increase the speed of processors while, at the same time, reducing the emission of heat and power consumption. 3D transistors have succeeded in manufacturing 22 nm processors, and this will enable the manufacturing of even smaller and lighter computers. Thanks to this, organization will be able to procure computers that are small and thus economize on office space (Lazakidou, 2009). Optic computing has also enabled transmission of messages over long distances within a short time. Therefore, companies have taken advantage of faster communication to form virtual offices. This has enabled companies to solve costs such as office rent and insurance of office premises. Virtual offices have also helped companies to save travelling allowances, which they might have incurred if they had acquired real office premises. Strong communication has also helped companies to outsource some of their non-core activities. Most companies, especially in United States, have outsourced some of their information technology departments to firms in India and China. This is because they are able to acquire expert services from qualified professionals at a low price.

The reduction in size of transistors has enabled manufacturers to produce smart phones. Business executives are able to send emails to their employees when they are on business trips or vacations. In addition to this, some smart phones have applications, which managers can use to sign necessary documents online, even if they are far away from their offices. Increase in the processor speed and efficiency has facilitated electronic data processing in most companies (Zelekowitz, 2011). Customers can book for airline tickets through the Internet and, at the same time, check the status of their airline tickets. It has also facilitated online banking since consumers are able to transfer funds from one account to another through the internet. Moreover, consumers are able to check their bank balances or purchase goods from online sites such as Amazon. The accuracy of computers has also increased due to the development in the transistors technology. Companies use computers to forecast their future share prices or liquidity. This helps them to deal with all the risks that may result from uncertain events that they may face.

Possible Improvements in the Scope of Information Technology

As scientists research on ways of reducing the limits of Moore’s Law, they will succeed in improving the scope of information technology in several ways. Nanowire wire transistors will bring massive improvements to the scope of information technology. These transistors are small and thus they will reduce the costs of producing chips. This will reduce the overall costs of computers and thus reduce the overall costs of sending messages using computers. Nanowires have regular crystal structures and uniform electronic crystals. Thanks to this, manufactures using these transistors will be able to manufacture high performance electronic devices (Hubner & Becner, 2010). Scientists on information technology have predicted that nanowires will be used to manufacture ultra-sensitive sensors that have the capability of detecting single molecules. Nonowire transistors will also facilitate the manufacturing of smaller and thinner smart phones since they facilitate the production of compact computer devices. This is because this technology offers the possibility of manufacturing layered memory devices and sensing circuitry on top of two or more chips. Nanowire transistors will also improve information technology since they will help in the manufacture of high-powered applications that operate under high temperatures. They have a core-shell structure that confines electrons in their electron holes; thus, the flow of electric currents is not affected by external pressure.

Parallel computing also promises to improve artificial intelligence. Information processing tasks in artificial intelligence require diverse computational needs that are provided by diverse computer systems. Parallel computing plays a crucial role since it provides a variety of architectural capabilities, therefore, allowing applications that have diverse execution requirements to be performed (Hubner & Beckner, 2010). Parallel computing has also brought massive developments in data mining technology. This is because it allows parallel programming, and thus section of a program can be used in looking for details in certain data while the other sections of the programs update any stored data. Researchers have also established that parallel computing could enable data mining programs to process about 100 rows of database per cycle. Parallel computing also promises to improve medical imaging and diagnosis. By means of using parallel computing, medical devices can accelerate the algorithm of brain fiber tracking. Complex structures, which are present in the human brain, are reconstructed by means of using noninvasive diffusion weighted imaging. Through parallel computing, the algorithm utilizes parallel structures, which are found in graphic processing units, and combines them with CUDA platform to increase the speed of fiber tracking devices.

The manufacture of 3D transistors promises to bring a revolution in information technology. After Intel produced the first 3D transistor, other manufacturers have started exploring how the 3 D transistors could be improved further. Soitec, a company that manufactures wafers, announced that it is currently exploring the options of producing wafers by utilizing fewer steps required in producing transistors, by using 3 D transistors technology (Zelkowitz, 2011). They will achieve this by making the channel of an ordinary planar transistor thin instead of wrapping gate around the three sides of a 3D channel. This will enable to eliminate the flow of free charges, therefore, lower the leak of currents. This will help in ensuring that the wafers are energy efficient.

Conclusion

In conclusion, the limitations of Moore’s law have led to the development of new technology of silicon transistors. Scientists in Intel produced 3D transistors that increased the processing speed and energy efficiency in Ivy Bridge 22-nanometers processors. In addition to this, it led to the manufacture of nanowire transistors that have made possible for engineers to produce reprogrammable circuits. Thanks to using nanowire transistors, scientists are able to develop low-power sensors that enable the processors to collect data more efficiently. Parallel computing also resulted from the limitations of Moore’s Law. Parallel computing enables processing cores to work together in parallel while solving computing instructions simultaneously. However, the above technological applications will face stiff completion from quantum computers, molecular computers and optical computers. This is because these computers do not use silicon, and thus they are able to represent data in more than binary digits. In addition to this, transistors produced using these forms of technology are smaller and consume less power than silicon transistors. The competition will benefit businesses since it will result in faster and more efficient computers. The scope of information technology will improve since the cost of communication will reduce. I feel that Moore’s Law will not succeed in the future since the quantum computers and molecular computers will be more efficient than silicon computers.