CSC/ECE 506 Spring 2013/1d mv

What is Moore's law?

Moore's law, named after Gordon Moore, co-founder of Intel, states that the number of transistors that can be placed on an integrated circuit will double approximately every two years<ref>http://www.computerhistory.org/semiconductor/timeline/1965-Moore.html</ref>. The original prediction in 1965 stated a doubling every 12 months, but in 1975, after microprocessors were introduced that were less dense he slowed the rate of doubling to it's current state of two years <ref>http://arstechnica.com/hardware/news/2008/09/moore.ars</ref>. Instead of giving an empirical formula predicting the rate of increase, Moore instead used prose, graphs, and images to convey these predictions and observations to the masses. This in some ways increased the staying power of Moore's law, allowing the industry to use it as a benchmark of success and a measurable determination of their success. Virtually all digital devices are in some way fundamentally linked to the growth set in place by Moore's law.<ref>http://en.wikipedia.org/wiki/Moore's_law</ref>

Historical context

One of the first reviews of progress in the semiconductor industry was written by G. Moore, for the 35th anniversary issue of Electronics magazine in 1965 <ref> http://download.intel.com/museum/Moores_Law/Articles-Press_Releases/Gordon_Moore_1965_Article.pdf </ref>. After only six years since the introduction of the first commercial planar transistor in 1959, Moore observed an astounding trend: the number of components per chip was roughly doubling every year, reaching about 64 components in 1965. Extrapolating this trend for a decade, Moore predicted that chips with 65,000 components would be available by 1975. In 1975, Moore revisited his 1965 prediction and provided some critical insights into the technological drivers of the observed trends <ref>http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1478174</ref>. Moore divided the advances in circuit complexity among its three principle technical drivers: increasing chip area, decreasing feature size, and improved device and circuit designs. Minimum feature sizes were decreasing by about 10% per year (resulting in transistors that were about 21% smaller in area, and an increase in transistors per area of 25% each year). Chip area was increasing by about 20% each year. These two factors alone resulted in a 50% increase in the number of transistors per chip each year. Design cleverness (eliminating nonfunctional chip area) made up the rest of the improvement (33%). By breaking the density improvement into its three technology drivers, Moore was able to extrapolate each trend into the future and predict a change in the slope of his observation. Moore saw the progress in lithography allowing continued feature size shrinks to one micrometer or less. Continued reductions in defect density and increases in wafer size would allow the die area trend to continue. Although improvements in device isolation and the development of the metal-oxide-semiconductor transistor had contributed to greater packing density, Moore saw the latest circuits as approaching their design limits. Seeing an end to the design cleverness trend in four or five years, Moore predicted a change in the slope of his trend from doubling every year, to approximately doubling every 18 months. Beginning in 1979, the industry followed the new Moore’s Law trend throughout the 1980s and early-1990s.

By the late-1970s, another interesting trend emerged: the bifurcation of Moore’s law into memory and logic trend lines. While memory chips continued to advance at the Moore’s law complexity limit, logic chips (such as microprocessors) advanced at a slower pace <ref>G. E. Moore, “VLSI: Some fundamental challenges,” IEEE Spectrum, vol. 16, no. 4, pp. 30–37, Apr. 1979. </ref> . In essence, the ability to put more transistors on a chip outstripped the ability to design chips with that many transistors for a market need. This split of Moore’s Law in two persisted for the next two decades, where by the year 2000 memory chips used 1 billion transistors, but advanced microprocessors had only 20–40 million transistors. While memory chips increased their transistors per chip by 1.58X per year, microprocessors saw only a 1.38X per year increase <ref>http://www.icknowledge.com/products/reports.html</ref>.

Semiconductor devices in computer hardware

Computer hardware has three basic components - a processor that rapidly executes instructions, memory that stores vast quantities of instructions and data (for instructions to manipulate), and input/output devices (like keyboards, displays, and interfaces to networks) that connect a computer to the outside world. Hardware designers seek to make processors execute instructions faster, memory provide information faster, and input/output devices that communicate faster. These electronic components are implemented on a semiconductor-based integrated circuit.

A semiconductor is a solid-state electronic device which can be switched to conduct or block electric currents <ref>http://research.microsoft.com/en-us/um/people/gray/moore_law.html</ref>. Semiconductors can be divided into two main groups: integrated circuits and discrete devices. Integrated circuits (ICs) consist of many active and passive elements that are fabricated and connected together in a single chip. ICs can be further broken down into digital or linear (analog) devices. Digital ICs store and process information expressed in binary numbers or "bits". They perform arithmetical operations or logical functions by manipulating binary signals.

Today, it is most common to implement a processor on a single silicon chip called a microprocessor. A microprocessor is an IC which provides on one chip functions equivalent to those contained in the processing unit of a computer. A microcontroller is a microprocessor and memory integrated on the same chip and is used in dedicated applications such as traffic signals, laser printers, or antilock brakes in automobiles. Microperipheral devices accompany microprocessors to handle a computer's related function such as graphics. Logic chips handle the mathematical treatment of formal logic by translating AND, OR, and NOT functions into a switching circuit, or gate. This category includes application-specific ICs (ASICs) including gate arrays, standard cells, and programmable logic devices (PLDs). Most computers today employ a single microprocessor, but large computers use several microprocessors (and, in a few cases, thousands).

Semiconductor memory devices store information in the form of electrical charges. Memory devices can be subdivided into volatile and non-volatile families. The Random-Access Memory (RAM) is a volatile memory product, which means it will lose stored information (charges) once power is turned off. This product sub-category includes the popular DRAM (Dynamic RAM) and, to a lesser degree, the SRAM (Static RAM). In contrast, non-volatile memory products retain information even after power is turned off. Non-volatile memory includes the ROM (Read Only Memory) and its "erasable" derivatives, the EPROM (Erasable Programmable ROM) and EEPROM (Electrically EPROM). In an EPROM, stored information is erased by exposure to ultraviolet light, whereas the EEPROM has the convenience of selective erasure of information through electrical impulses rather than exposure to ultraviolet light. "Flash Memory" is an IC which has the ability to bulk erase its entire contents simultaneously. It shares the advantage of other non-volatile memory in that it retains information when power is turned off. Current desktop computer memories are implemented in several chips and backed up by one or more magnetic disks. Larger computers can employ hundreds of memory chips and disks.

Manufacturing techniques

At the heart of Moore's Law is the transistor. Computer chips contain hundreds of millions of transistors embedded on a wafer of silicone. To make these chips, first a “stencil” is made containing the outlines of millions of transistors. The stencil is then placed over a silicone wafer which is sensitive to ultraviolet light. The light penetrates through the gaps of the stencil and exposes the silicon wafer, which is then is bathed in acid, carving the outlines of the circuits and the design of millions of transistors. Since the wafer consists of many conducting and semiconducting layers, the acid cuts into the wafer at different depths and patterns, so one can create circuits of enormous complexity.

One reason why Moore’s law has relentlessly increased the power of chips is because UV light can be tuned so that its wavelength is smaller and smaller, making it possible to etch increasingly tiny transistors onto silicon wafers. Since UV light has a wavelength as small as 10 nanometers, the smallest transistor that can be etched is about thirty atoms across.<ref>http://en.wikipedia.org/wiki/Photolithography</ref> Due to physical limitations, This process cannot go on forever. At some point, it will be physically impossible to etch smaller effective transistors, and Moore’s law as we understand it will finally collapse.

Currently, estimates predict that around 2020 or soon afterward, Moore’s law will gradually cease to hold true for traditional transistor technology. Transistors will be so small that quantum effects will begin to take over and electrons will "leak" out of the wires.<ref>http://computer.howstuffworks.com/small-cpu2.htm</ref><ref>http://www.monolithic3d.com/2/post/2011/09/is-there-a-fundamental-limit-to-miniaturizing-cmos-transistors1.html</ref> For example, the thinnest layer inside a computer will be about five atoms across. At that size, quantum effects will become dominant and transistors will not function as they currently do without other technological advances.

A common misconception

Moore's Law is often linked to performance improvements as measured in CPU clock speeds. In the 1980's, former Intel executive David House stated that chip performance would double every 18 months.<ref>http://news.cnet.com/Myths-of-Moores-Law/2010-1071_3-1014887.html</ref> This is a consequence of Moore's Law, but it is not what Moore's Law actually claims. In fact, due to heat dissipation issues<ref>http://www.gotw.ca/publications/concurrency-ddj.htm</ref><ref>http://techtalk.pcpitstop.com/2007/08/06/cpu-clock-speeds/</ref>, performance as measured in clock speed has remained flat since 2005<ref>http://www.kmeme.com/2010/09/clock-speed-wall.html</ref> while the number of transistors continues to double roughly every 2 years.

Moore's law, past to present

Reviewing data from the inception of Moore's law to the present shows that, consistent to Moore's prediction, the number of transistors on a chip has doubled approximately every 2 years. There are several contributing factors, that had they not been developed, could have slowed or plateaued Moore's law. One of these is the invention Dynamic random access memory (DRAM). This is a type of random access memory that allows for the storage of each bit in a separate capacitor on an integrated circuit. The main advantage of DRAM over its predecessor, SRAM, is that only one transistor and a capacitor are required per bit, compared to four or six transistors with SRAM. Another is most certainly the complementary metal-oxide-semiconductor (CMOS). This allowed a higher density of logic functions on a chip with the added benefit of low power consumption and electrical noise immunity. Lastly was the invention of the integrated circuit itself. Moore's law isn't only responsible for making larger and faster chips, but also smaller, cheaper, and more efficient ones as well.

As visible in the examples below and to the right, Moore's law has seemed to hold at an overall constant rate of growth consistent to what Moore predicted. There is some variation from year to year that can be explained by the introduction of new technology, manufacturing or otherwise that helped kick-start Moore's law back on track. This is most evident with the "Dual-Core Itanium 2" processor. It was ahead of it's competitors by a full four years. Looking at the example to the right, we can visibly see a dip in transistor count during 1995-2003. But this is balanced out by an almost equal increase in the following eight years.

Processor	Transistor count	Date of introduction	Manufacturer	Process	Area
Intel 4004	2,300	1971	Intel	10 µm	12 mm²
Intel 8008	3,500	1972	Intel	10 µm	14 mm²
MOS Technology 6502	3,510	1975	MOS Technology		21 mm²
Motorola 6800	4,100	1974	Motorola		16 mm²
Intel 8080	4,500	1974	Intel	6 μm	20 mm²
RCA 1802	5,000	1974	RCA	5 μm	27 mm²
Intel 8085	6,500	1976	Intel	3 μm	20 mm²
Zilog Z80	8,500	1976	Zilog	4 μm	18 mm²
Motorola 6809	9,000	1978	Motorola	5 μm	21 mm²
Intel 8086	29,000	1978	Intel	3 μm	33 mm²
Intel 8088	29,000	1979	Intel	3 μm	33 mm²
Intel 80186	55,000	1982	Intel
Motorola 68000	68,000	1979	Motorola	4 μm	44 mm²
Intel 80286	134,000	1982	Intel	1.5 µm	49 mm²
Intel 80386	275,000	1985	Intel	1.5 µm	104 mm²
Intel 80486	1,180,000	1989	Intel	1 µm	160 mm²
Pentium	3,100,000	1993	Intel	0.8 µm	294 mm²
AMD K5	4,300,000	1996	AMD	0.5 µm
Pentium II	7,500,000	1997	Intel	0.35 µm	195 mm²
AMD K6	8,800,000	1997	AMD	0.35 µm
Pentium III	9,500,000	1999	Intel	0.25 µm
AMD K6-III	21,300,000	1999	AMD	0.25 µm
AMD K7	22,000,000	1999	AMD	0.25 µm
Pentium 4	42,000,000	2000	Intel	180 nm
Atom	47,000,000	2008	Intel	45 nm
Barton	54,300,000	2003	AMD	130 nm
AMD K8	105,900,000	2003	AMD	130 nm
Itanium 2	220,000,000	2003	Intel	130 nm
Cell	241,000,000	2006	Sony/IBM/Toshiba	90 nm
Core 2 Duo	291,000,000	2006	Intel	65 nm
AMD K10	463,000,000	2007	AMD	65 nm
AMD K10	758,000,000	2008	AMD	45 nm
Itanium 2 with 9MB cache	592,000,000	2004	Intel	130 nm
Core i7 (Quad)	731,000,000	2008	Intel	45 nm	263 mm²
Six-Core Xeon 7400	1,900,000,000	2008	Intel	45 nm
POWER6	789,000,000	2007	IBM	65 nm	341 mm²
Six-Core Opteron 2400	904,000,000	2009	AMD	45 nm	346 mm²
16-Core SPARC T3	1,000,000,000	2010	Sun/Oracle Corporation|Oracle	40 nm	377 mm²
Core i7 (Gulftown)	1,170,000,000	2010	Intel	32 nm	240 mm²
8-core POWER7	1,200,000,000	2010	IBM	45 nm	567 mm²
z196	1,400,000,000	2010	IBM	45 nm	512 mm²
Dual-Core Itanium 2	1,700,000,000	2006	Intel	90 nm	596 mm²
Tukwila	2,000,000,000	2010	Intel	65 nm	699 mm²
Core i7 (Sandy Bridge-E)	2,270,000,000	2011	Intel	32 nm	434 mm²
Nehalem-EX	2,300,000,000	2010	Intel	45 nm	684 mm²
10-Core Xeon Westmere-EX	2,600,000,000	2011	Intel	32 nm	512 mm²

The lesser known second law

There is another fundamental principle known as Rock's law, this law is a direct consequence to Moore's law in that the cost to produce transistors on a chip may go down, these costs instead flow towards manufacturing, testing, and research and development. The law states that the cost of a semiconductor chip fabrication plant doubles ever four years. Simply put, in order for Moore's law to hold, Rock's law must also hold.<ref>http://en.wikipedia.org/wiki/Rock%27s_law</ref>

Due to current economic downturns, Moore's law may see troubling times ahead. Due to the increase investment towards research and development and fabrication cost, it may soon be difficult to meet an adequate return on investment given current production standards. New and innovative methods will have to be developed.<ref>http://ieeexplore.ieee.org.prox.lib.ncsu.edu/stamp/stamp.jsp?tp=&arnumber=5227172</ref>

Beyond Moore's Law

There are a few new technologies that have the potential to change the underlying architecture of processors and extend performance gains past the theoretical limits of traditional transistors.

Does transistor count matter?

Moore's law concerns only the doubling of transistors on the same die space every 2 years. While some of these new technologies deal directly with adding more transistors into the same amount of space, others take a different approach to boost overall computational performance. While not strictly following Moore's law, per se, these advanced designs will lead to a continuation of the increase in computational power that can be harnessed from hardware. They are included in the discussion to illustrate that performance is not necessarily dependent on the number of transistors that can be placed on a die. Novel approaches, such as 3-D transistor manufacturing will allow for greater densities, but other approaches, such as quantum computing operate in a different way than the traditional transistor to solve the same problem more efficiently.<ref>http://www.monolithic3d.com/2/post/2011/09/is-there-a-fundamental-limit-to-miniaturizing-cmos-transistors1.html</ref><ref>http://www.iue.tuwien.ac.at/phd/wittmann/node6.html</ref>

The Memristor

Currently being developed by Hewlett Packard, the memristor is a new type of transistor that combines both electrical charge and magnetic flux. As current flows in one direction through the circuit, resistance increases. Reversing the flow will decrease the resistance and stopping the flow will leave the resistance in the current state. This type of structure allows for both data storage and data processing (logic gate construction). Currently, it is postulated that memristors could be layered in three dimensions on silicone, yielding data and transistor densities of up to 1000 times greater than currently available. HP has reported the ability to fit 100GB in a square centimeter<ref>http://www.eetimes.com/electronics-news/4076910/-Missing-link-memristor-created-Rewrite-the-textbooks</ref> and with the ability to layer memristors, this could lead to pocket devices with a capacity of over 1 petabyte.

Some advanced theoretical capabilities of memristors are the ability to store more than one state, which can lead to analog computing. Memristor technology may also provide an excellent architecture for synaptic modeling and self-learning systems.

Quantum computing

Quantum computing works by essentially allowing all available bits to enter into superposition. Using this superposition, each "q-bit" can be entangled with other q-bits to represent multiple states at once. By using quantum logic gates, the qbits can be manipulated to find the desired state among the superposition of states. This has great potential for drastically shortening the time necessary to solve several important problems, including integer factorization and the discrete logarithm problem, upon which much current encryption is based. Quantum computing faces several technical issues, including decoherence, which makes quantum computers difficult to construct and maintain.

Ballistic Deflection Transistors

Another promising avenue is a re-design of the traditional transistor. Essentially, single electrons are passed through a transistor and deflected into one path or the other, thus delivering a 0 or a 1. The theoretical speed of these transistors is in the terahertz range.<ref>http://www.rochester.edu/news/show.php?id=2585</ref>

Other technologies

The arena of research to produce an alternative to the traditional transistor includes many novel approaches. They include (but are not limited to):

• Optical Computing
• DNA Computing
• Molecular Electronics
• Spintronics
• Chemical Computing
• Artificial Neural Networks
• Unconventional Computing

Conclusions

The demise of Moore's Law has been predicted several times during the past 40 years, but transistor counts continue to follow a two year doubling on average. With the traditional transistor approach, inevitable physical limits will be reached around the 16 nm process, due to quantum tunneling<ref>http://news.cnet.com/2100-1008-5112061.html</ref>. If this is true, the current pace of innovation would lead to hitting "Moore's Wall" around 2022, or in about 10 years. This "10 year horizon" for Moore's Law has existed since the early 1990's, with new designs, processes, and breakthroughs which continue to extend the timeline.<ref>http://arxiv.org/pdf/astro-ph/0404510v2.pdf</ref><ref>http://java.sys-con.com/node/557154</ref> New technologies that leverage three dimensional chip architecture would allow for years of continued growth in transistor counts and exotic designs could further increase the theoretical capacity of transistors in a particular space. If the past is used as a predictor for future trends, it is safe to say that the end of Moore's Law "is about 10 years away".

On another note, if we relax the definition of Moore's Law to include computational performance gains, we open a whole new avenue by which to measure computing power. Most of the easy gains in performance related to transistor counts have been realized, but new designs of how basic computing is performed can theoretically yield large increases in performance without doubling of transistor counts or extremely high power requirements.<ref>http://abcnews.go.com/Technology/story?id=4006166&page=1#.TzAbDcVA_H4</ref><ref>http://www.gotw.ca/publications/concurrency-ddj.htm</ref> The era of the traditional transistor is not quite over yet, but the relevance of transistor counts may be nearing it's end.

References

<references/>