imported>Gareth Leng |
|
| (45 intermediate revisions by 10 users not shown) |
| Line 1: |
Line 1: |
| [[Image:Moore_Law_diagram_(2004).jpg|thumb|350px|Growth of [[transistor count]]s for [[Intel]] processors (dots) and Moore's Law (upper line=18 months; lower line=24 months)]] | | {{subpages}} |
| | {{TOC|Right}} |
| | <!--[[Image:Moore_Law_diagram_(2004).jpg|thumb|350px|Growth of [[Electronic switch#Transistor|transistor counts]] for [[Intel]] processors (dots) and Moore's Law (upper line, 18 months; lower line, 24 months)]]--> |
|
| |
|
| '''Moore's Law''' is the [[empirical]] observation that the [[transistor count|transistor density]] of [[integrated circuit]]s, with respect to minimum component cost, doubles every 24 months<ref name="IntelInterview">{{cite web| year =2005|url=ftp://download.intel.com/museum/Moores_Law/Video-Transcripts/Excepts_A_Conversation_with_Gordon_Moore.pdf| title =Excerpts from A Conversation with Gordon Moore: Moore’s Law| format ={{PDFlink}}| pages =1| publisher=[[Intel|Intel Corporation]]| accessdate =May 2| accessyear =2006}}</ref>. It is attributed to [[Gordon Moore|Gordon E. Moore]]<ref>Not to be confused with another ''G.E. Moore'', the [[philosopher]] [[George Edward Moore]], the creator of [[Moore's paradox]].</ref>, a co-founder of [[Intel]]. Moore's statement is in his publication "Cramming more components onto [[integrated circuits]]", ''[[Electronics (magazine)|Electronics Magazine]]'' 19 April 1965<ref name="IntelInterview">{{cite web| year =2005|url=ftp://download.intel.com/museum/Moores_Law/Video-Transcripts/Excepts_A_Conversation_with_Gordon_Moore.pdf| title =Excerpts from A Conversation with Gordon Moore: Moore’s Law| format =PDF| pages =1| publisher=[[Intel|Intel Corporation]]| accessdate =May 2| accessyear =2006}}</ref>: | | '''Moore's law''' is the prediction that the [[Electronic switch#Transistors|transistor]] density inside [[integrated circuit]]s will double every two years, and that prices will decline at the same time<ref name="IntelInterview">{{cite web| year =2005|url=ftp://download.intel.com/museum/Moores_Law/Video-Transcripts/Excepts_A_Conversation_with_Gordon_Moore.pdf| title =Excerpts from A Conversation with Gordon Moore: Moore’s Law| format =PDF| pages =1| publisher=[[Intel|Intel Corporation]]| accessdate =May 2| accessyear =2006}}</ref>. |
| {{cquote|The complexity for minimum component costs has increased at a rate of roughly a factor of two per year ... Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years. That means by [[1975]], the number of components per integrated circuit for minimum cost will be 65,000. I believe that such a large circuit can be built on a single wafer.}}
| | The phenomemon predicted by Moore's law, first described in 1965, has held remarkably true to date, and experts predict that this trend might continue until ~2020 or so, declining at the point where [[Electronic switch|switching element]] sizes reach the molecular level. Moore's law is not really a law, but rather more a “rule of thumb” or a practical way to think about something. |
|
| |
|
| Under the assumption that chip 'complexity' is proportional to the number of [[transistor]]s, regardless of what they do, the law has largely held true to date. However, as the per-transistor complexity is less in large [[random access memory|RAM]] cache arrays than in [[execution units]], the validity of Moore's Law may be more questionable. Moore's observation was named a "law" by the [[Caltech]] professor and [[very-large-scale integration|VLSI]] pioneer [[Carver Mead]]<ref name="IntelInterview" />.
| | Another way to think about it is that density is inversely proportional to the distance that signals must traverse, so Moore's Law does not only address the computational capability of more circuit elements, but workarounds to speed-of-light limitations. |
| | {{quotation|186,300 miles per second. It's not just a good idea. It's the Law.|Seen at a T-shirt at an IETF meeting}} |
|
| |
|
| Moore may have heard [[Douglas Engelbart]], a co-[[inventor]] of the mechanical [[computer mouse]], discuss the projected downscaling of integrated circuit size in a [[1960]] lecture.<ref>[http://theory.kitp.ucsb.edu/~paxton/doug.html NY Times article] April 17 2005</ref> In 1975, Moore projected a doubling only every two years. He is adamant that he never said "every 18 months", but that is how it has been quoted. The [[SEMATECH]] roadmap follows a 24 month cycle. In April 2005, Intel offered $10,000 for a copy of the original ''[[Electronics (magazine)|Electronics Magazine]]''. <ref>{{cite web| year =2005|url=http://news.zdnet.co.uk/0,39020330,39194694,00.htm| title =$10,000 reward for Moore's Law original |date=2005-04-12| author=Michael Kanellos|publisher=CNET News.com | accessdate =June 24| accessyear =2006}}</ref> | | Moore's law is named for [[Gordon Moore]], a co-founder of [[Intel]], who wrote about it in "Cramming more components onto [[integrated circuits]]", ''[[Electronics (magazine)|Electronics Magazine]]'' 19 April 1965<ref name="IntelInterview">{{cite web| year =2005|url=ftp://download.intel.com/museum/Moores_Law/Video-Transcripts/Excepts_A_Conversation_with_Gordon_Moore.pdf| title =Excerpts from A Conversation with Gordon Moore: Moore’s Law| format =PDF| pages =1| publisher=[[Intel|Intel Corporation]]| accessdate =May 2| accessyear =2006}}</ref>: |
| | <blockquote>''The complexity for minimum component costs has increased at a rate of roughly a factor of two per year ... Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years. That means by 1975, the number of components per integrated circuit for minimum cost will be 65,000. I believe that such a large circuit can be built on a single wafer''</blockquote> |
|
| |
|
| ==Formulations of Moore's Law==
| | Although named for him, Gordon Moore may not have invented Moore's law; instead, he may have heard [[Douglas Engelbart]], a co-[[inventor]] of the mechanical [[computer mouse]], discuss the projected downscaling of [[integrated circuit]] size in a 1960 lecture.<ref>[http://theory.kitp.ucsb.edu/~paxton/doug.html NY Times article] April 17 2005</ref>. Moore's observation was named a 'law' by the [[Caltech]] professor and [[VLSI]] pioneer [[Carver Mead]]<ref name="IntelInterview" />. |
| The most popular formulation is of the doubling of the number of [[transistor]]s on [[integrated circuit]]s (a rough measure of computer processing power) every 18 months. At the end of the 1970s, Moore's Law became known as the limit for the number of transistors on the most complex chips. However, it is also common to cite Moore's Law to refer to the rapidly continuing advance in computing power per [[unit cost]]. A similar law has held for [[hard disk]] storage cost per unit of information. The rate of progression in [[disk storage]] over the past decades has actually sped up more than once, corresponding to the utilization of [[error correcting code]]s, the [[magnetoresistive effect]] and the [[giant magnetoresistive effect]]. The current rate of increase in [[hard drive]] capacity is roughly similar to the rate of increase in transistor count and has been dubbed [[Kryder's Law]]. However, recent trends show that this rate is dropping, and has not been met for the last three years. Another version states that [[Random Access Memory|RAM]] storage capacity increases at the same rate as processing power. However, memory speeds have not increased as fast as [[CPU]] speeds in recent years, leading to a heavy reliance on caching in current computer systems.
| |
|
| |
|
| ==An industry driver==
| | In 1975, Moore projected a doubling only every two years. He is adamant that he never said "every 18 months", but that is how it has been quoted. The [[SEMATECH]] roadmap follows a 24 month cycle. In April 2005, Intel offered $10,000 for a copy of the original ''[[Electronics (magazine)|Electronics Magazine]]''. <ref>{{cite web| year =2005|url=http://news.zdnet.co.uk/0,39020330,39194694,00.htm| title =$10,000 reward for Moore's Law original |date=2005-04-12| author=Michael Kanellos|publisher=CNET News.com | accessdate =June 24| accessyear =2006}}</ref> |
| Although Moore's Law was initially an observation and forecast, the more widely it became accepted, the more it served as a goal for an entire industry. This drove both [[marketing]] and [[engineering]] departments of [[semiconductor]] manufacturers to focus enormous energy aiming for the increase in processing power that it was presumed one or more of their competitors would soon attain. In this regard, it can be viewed as a [[self-fulfilling prophecy]]. The implications of Moore's Law for computer component suppliers are significant. A typical major design project (such as an all-new CPU or hard drive) takes 2-5 years to reach production-ready status. In consequence, component manufacturers face enormous timescale pressures—just a few weeks of delay can mean the difference between success and massive losses. Expressed as "a doubling every 18 months", Moore's Law suggests the phenomenal progress of technology in recent years. Expressed on a shorter [[timescale]] however, Moore's Law equates to an average performance improvement in the industry as a whole of over 1% ''per week''. For a manufacturer in the competitive CPU market, a new product that is expected to take three years to develop and is just two or three months late is 10 to 15% slower, bulkier, or lower in storage capacity than competing products, and is usually unsellable.
| | |
| | It has become common practice to cite Moore's Law as a predictor for the rapid advance in computing power per [[unit cost]] in a variety of computer-related technologies, such as [[Hard Disk|hard disk]] storage cost per unit of information, even though such advances may have little to do with transistor technology advances. Moore's Law has become synonymous with saying "things very quickly gain capabilities, while cost falls". |
| | ==Processor speed not always the limiting factor== |
| | A new important chapter started for Moore's law in 2003. The cycles per second for microprocessors went flat, and processor designers began to keep Moore's law alive by adding |
| | extra cores. [[multicore]] microprocessors in personal computers. <ref Name=ConRev>[http://www.ddj.com/cpp/184401916 Multi-Core]</ref> |
| | ===Power consumption=== |
| | The problem of absolute power consumption surfaced at about the same time. |
| | The power consumption of computers is now a non-trivial fraction of each modern nation's total power output. (about 2% in 2005) The current limitations to Moore's law are not due to hardware engineering, rather the burden is on programmers to make use of multiple cores and the burden is on society to pay the cost of the extra electrical load. In response, metrics like flops/watt are being used to help keep computation affordable. Note: 1.2% of total US electrical output in 2005 went for servers alone. http://dl.klima2008.net/ccsl/koomey_long.pdf (Koomey, 2008) |
| | ===Processor faster; software slower=== |
| | There is a joke about "Gates' Law", a sort of inverse of Moore's Law: |
|
| |
|
| ==Future trends==
| | {{quotation| “The speed of software halves every 18 months.” This oft-cited law is an ironic comment on the tendency of software bloat to outpace the every-18-month doubling in hardware capacity per dollar predicted by Moore's Law. The reference is to Bill Gates; Microsoft is widely considered among the worst if not the worst of the perpetrators of bloat. [http://www.catb.org/~esr/jargon/html/G/Gatess-Law.html Jargon file]}} |
| As of 2006, current PC processors are fabricated at the 90 [[Nanometre|nm]] level and 65nm chips are just being rolled out by Intel ([[Pentium D]] & [[Intel Core]]). A decade ago, chips were built at a 500 nm level. Companies are working on using [[nanotechnology]] to solve the complex engineering problems involved in producing chips at the 45nm, 30nm, and even smaller levels—a process that will postpone the industry meeting the limits of Moore's Law. Recent computer [http://public.itrs.net/ industry technology "roadmaps"] predict (as of 2001) that Moore's Law will continue for several chip generations. Depending on the doubling time used in the calculations, this could mean up to 100 fold increase in transistor counts on a chip in a decade. The semiconductor industry technology roadmap uses a three-year doubling time for [[microprocessor]]s, leading to about nine-fold increase in a decade. In early 2006, [[IBM]] researchers announced that they had developed a technique to print circuitry only 29.9nm wide using [[Ultraviolet|deep-ultraviolet]] (DUV, 193-nanometer) [[Photolithography|optical lithography]]. IBM claims that this technique may allow chipmakers to use current methods for seven years while continuing to achieve results predicted by Moore's Law. New methods that can achieve smaller circuits are predicted to be substantially more expensive.
| | ===Memory=== |
| | Some applications are memory, not processor limited. Admittedly, memory is also dependent on many of the same semiconductor technologies as processors, but a given computer may not be able to accept more physical memory. Desktop computers with Microsoft operating systems often improve more with more memory than a replacement faster processor -- although the situation can be confusing if a different processor is needed to accept more memory. |
|
| |
|
| Since the rapid [[exponential growth|exponential improvement]] could (in theory) put 100 GHz personal computers in every home and 20 GHz devices in every pocket, some commentators have speculated that sooner or later computers will meet or exceed any conceivable need for [[computation]]. This is only true for some problems—there are others where exponential increases in processing power are matched or exceeded by exponential increases in complexity as the problem size increases. See [[computational complexity theory]] and [[complexity classes P and NP]] for a discussion of such problems, which occur very commonly in applications such as [[Scheduling (computing)|scheduling]].
| | In the early 1990s, the crisis in Internet routing, which was worked around with [[Classless Inter-Domain Routing]] (CIDR), was a practical consequence of the Cisco AGS, then the most common Internet core router, holding a maximum of 16 megabytes of memory. Until the CIDR techniques went into operational use, the number of routes was doubling every five months and threatening to overload the routing tables of those routers. |
| | | ===Networks=== |
| The exponential increase in frequency of operation as the only method of increasing computation speed is misleading. What matters is the exponential increase in useful work (or instructions) executed per unit time. In fact, newer processors are actually being made at lower clock speeds, with focus on larger [[cache]]s and multiple computing cores. The reason for this is that higher clock speeds correspond to exponential increases in temperature, such that it becomes almost impossible to produce a CPU that runs reliably at speeds faster than about 4.3 GHz.
| | Another set of laws deals with the growth of connectivity in computer networks. Sarnoff's Law deals with the value of a one-to-many radio or television broadcast network, as opposed to Metcalfe's Law about any-to-any internets. |
| | | ==References== |
| [[Extrapolation]] partly based on Moore's Law has led [[futurologists]] such as [[Vernor Vinge]], [[Bruce Sterling]] and [[Ray Kurzweil]] to speculate about a [[technological singularity]]. However, in April 2005, Gordon Moore himself stated in an interview that the law may not hold for too long, as transistors might reach the limits of miniaturization at atomic levels.
| | {{reflist|2}}[[Category:Suggestion Bot Tag]] |
| | |
| {{cquote|In terms of size [of transistor] you can see that we're approaching the size of [[atom]]s which is a fundamental barrier, but it'll be two or three generations before we get that far—but that's as far out as we've ever been able to see. We have another 10 to 20 years before we reach a fundamental limit. By then they'll be able to make bigger chips and have transistor budgets in the billions.<ref>{{cite web| year =2005|url=http://www.techworld.com/opsys/news/index.cfm?NewsID=3477| title =Moore's Law is dead, says Gordon Moore|date=2005-04-13| author=Manek Dubash|publisher=Techworld | accessdate =June 24| accessyear =2006}}</ref>}}
| |
| | |
| While this time horizon for Moore's Law scaling is possible, it does not come without underlying engineering challenges. One of the major challenges in integrated circuits that use [[nanoscale]] transistors is increase in parameter variation and leakage currents. Because of variation and leakage, the design margins available to do predictive design is becoming harder, and such systems dissipate considerable power even when not switching. Adaptive and statistical design along with leakage power reduction is critical to sustain scaling of [[CMOS]]. <ref>[http://www.springer.com/sgw/cda/frontpage/0,11855,4-40109-22-52496396-0,00.html Leakage in Nanometer CMOS Technologies]</ref>. Other scaling challenges include:
| |
| # The ability to control [[parasitic]] [[resistance]] and [[capacitance]] in transistors,
| |
| # The ability to reduce resistance and capacitance in electrical [[interconnect]]s,
| |
| # The ability to maintain proper transistor [[electrostatics]] that allow the [[gate terminal]] to control the ON/OFF behavior,
| |
| # Increasing effect of line edge roughness,
| |
| # [[Dopant]] [[fluctuations]],
| |
| # System level power delivery,
| |
| # [[Thermal]] design to effectively handle the dissipation of delivered power, and
| |
| # Solve all these challenges with ever-reducing cost of manufacturing of the overall system.
| |
| | |
| [[Image:PPTMooresLawai.jpg|thumb|right|450px|Kurzweil expansion of Moore's Law shows that due to [[paradigm shift]]s the underlying trend holds true from [[integrated circuits]] to earlier [[transistor]]s, [[vacuum tube]]s, [[relay]]s and [[electromechanics|electromechanical]] computers.]]
| |
| | |
| [[Ray Kurzweil|Kurzweil]] projects that a continuation of Moore's Law until [[2019]] will result in transistor features just a few atoms in width. Although this means that the strategy of ever finer [[photolithography]] will have run its course, he speculates that this does not mean the end of Moore's Law:
| |
| | |
| {{cquote|Moore's Law of Integrated Circuits was not the first, but the fifth [[paradigm]] to provide accelerating price-performance. Computing devices have been consistently multiplying in power (per unit of time) from the mechanical calculating devices used in the [[U.S. Census, 1890|1890 US Census]], to [[Turing]]'s relay-based "Robinson" machine that cracked the [[Nazi]] [[enigma code]], to the [[CBS]] vacuum tube computer that predicted the election of [[Eisenhower]], to the transistor-based machines used in the first [[space launch]]es, to the integrated-circuit-based personal [computers].<ref>{{cite web|url=http://www.kurzweilai.net/articles/art0134.html?printable=1| title =The Law of Accelerating Returns|date=2001-03-07| author=Ray Kurzweil|publisher=KurzweilAI.net | accessdate =June 24| accessyear =2006}}</ref>}}
| |
| | |
| Thus, Kurzweil conjectures that it is likely that some new type of technology will replace current integrated-circuit technology, and that Moore's Law will hold true long after 2020. He believes that the [[exponential growth]] of Moore's Law will continue beyond the use of integrated circuits into technologies that will lead to the [[technological singularity]]. The [[Law of Accelerating Returns]] described by Ray Kurzweil has in many ways altered the public's perception of Moore's Law. It is a common (but mistaken) belief that Moore's Law makes predictions regarding all forms of technology, when it actually only concerns [[semiconductor]] [[circuit]]s. Many [[futurist]]s still use the term 'Moore's Law' to describe ideas like those of Kurzweil.
| |
| | |
| Krauss and Starkman announced an ultimate limit of around 600 years in their paper [http://arxiv.org/abs/astro-ph/0404510&e=10129 "Universal Limits of Computation"], based on rigorous estimation of total information-processing capacity of any system in the Universe. Then again, the law has often met obstacles that appeared insurmountable, before soon surmounting them. In that sense, Mr. Moore says he now sees his law as more beautiful than he had realised. "Moore's Law is a violation of [[Murphy's Law]]. Everything gets better and better." <ref>
| |
| {{cite web| year =2005|url=http://economist.com/displaystory.cfm?story_id=3798505| title =Moore's Law at 40 - Happy birthday|date=2005-03-23| publisher=The Economist| accessdate =June 24| accessyear =2006}}</ref>
| |
| | |
| ==Other considerations==
| |
| Not all aspects of [[computing technology]] develop in capacities and speed according to Moore's Law. [[Random Access Memory|Random Access Memory (RAM)]] speeds and hard drive seek times improve at best a few percentages per year. Since the capacity of RAM and hard drives is increasing much faster than is their access speed, intelligent use of their capacity becomes more and more important. It now makes sense in many cases to trade space for time, such as by precomputing indexes and storing them in ways that facilitate rapid access, at the cost of using more disk and memory space: space is getting cheaper relative to time. Another, sometimes misunderstood, point is that exponentially improved [[hardware]] does not necessarily imply exponentially improved [[software]] to go with it. The productivity of software developers most assuredly does not increase exponentially with the improvement in hardware, but by most measures has increased only slowly and fitfully over the decades. Software tends to get larger and more complicated over time, and [[Wirth's law]] even states that "Software gets slower faster than hardware gets faster". Moreover, there is [[popular misconception]] that the clock speed of a processor determines its speed, also known as the [[Megahertz Myth]]. This actually also depends on the number of instructions per tick which can be executed (as well as the complexity of each instruction, see [[Million instructions per second|MIPS]], [[RISC]] and [[CISC]]), and so the clock speed can only be used for comparison between two identical circuits. Of course, other factors must be taken into consideration such as the [[bus size]] and speed of the [[peripheral]]s. Therefore, most popular evaluations of "computer speed" are inherently biased, without an understanding of the underlying technology. This is especially true now that popular manufacturers play with public perception of speed, focusing on advertising the clock rate of new products. <ref>{{cite web| url=http://news.zdnet.co.uk/hardware/chips/0,39020354,2107456,00.htm | title =Intel, Aberdeen attack AMD speed ratings |date=2006-06-24| author=Matthew Broersma|publisher=ZDNet UK| accessdate =June 24| accessyear =2006}}</ref>
| |
| | |
| As the cost to the consumer of computer power falls, the cost for producers has the opposite trend: R&D, manufacturing, and test costs have increased steadily with each new generation of chips. As the cost of semiconductor equipment is expected to continue increasing, manufacturers must sell more and more chips to remain profitable. (The cost to tape-out a chip at 0.18 μm was roughly $300,000 USD. The cost to tape-out a chip at 90 nm exceeds $750,000 USD, and the cost is expected to exceed $1.0M USD for 65 nm) In recent years, analysts have observed a decline in the number of "design starts" at advanced process nodes (0.13 μm and below.) While these observations were made in the period after the 2000 economic downturn, the decline may be evidence that traditional manufacturers in the long-term [[global market]] cannot economically sustain Moore's Law. However, Intel was reported in 2005 as stating that the downsizing of [[silicon]] chips with good economics can continue for the next decade <ref>{{cite web| url=http://news.com.com/New+life+for+Moores+Law/2009-1006_3-5672485.html?tag=nl | title=New life for Moores Law |date=2006-04-19| |publisher=CNET News.com| accessdate =June 24| accessyear =2006}}</ref>. Intel's prediction of increasing use of materials other than silicon, was verified in mid-[[2006]], as was its intent of using trigate transistors around 2009. Researchers from [[IBM]] and [[Georgia Tech]] created a new speed record when they ran a silicon/[[germanium]] [[helium]] [[supercooled]] chip at 500 GHz <ref>{{cite web| url=http://news.bbc.co.uk/1/hi/technology/5099584.stm | title =Chilly chip shatters speed record |date=2006-06-20| |publisher=BBC Online| accessdate =June 24| accessyear =2006}}</ref>. The chip operated above 500 GHz at 4.5 [[Kelvin|K]] (451 degrees below zero Fahrenheit) <ref>{{cite web| url=http://www.gatech.edu/news-room/release.php?id=1019 | title =Georgia Tech/IBM Announce New Chip Speed Record |date=2006-06-20| |publisher=Georgia Institute of Technology| accessdate =June 24| accessyear =2006}}</ref> and simulations showed that it could likely run at 1 THz (1,000 GHz).
| |
| | |
| ==References and notes==
| |
| <div class="references-small"><references/></div>
| |
| | |
| ==External links==
| |
| ===Articles===
| |
| * [http://www.intel.com/technology/mooreslaw/index.htm Intel's information page on Moore's Law] – With link to Moore's original 1965 paper
| |
| * [http://www.intel.com/pressroom/kits/events/moores_law_40th/index.htm Intel press kit] released for Moore's Law's 40th anniversary, with a [ftp://download.intel.com/pressroom/images/events/moores_law_40th/Moores_Law_Original_Graph.jpg 1965 sketch] by Moore
| |
| * [http://firstmonday.org/issues/issue7_11/tuomi/index.html The Lives and Death of Moore's Law] – By [[Ilkka Tuomi]]; a detailed study on Moore's Law and its historical evolution and [http://www.kurzweilai.net/meme/frame.html?main=/articles/art0593.html its criticism] by Kurzweil.
| |
| * [http://news.com.com/2100-1006_3-5607422.html Moore says nanoelectronics face tough challenges] – By Michael Kanellos, CNET News.com, 9 March 2005
| |
| * [http://home.earthlink.net/~moores-law/ Blog and news on End of Moore's Law] – Moore's Law graph showing estimated end time, other related graphics
| |
| * [http://www.nytimes.com/2005/04/18/technology/18moore.html It's Moore's Law, But Another Had The Idea First] by [[John Markoff]]
| |
| * [http://news.bbc.co.uk/2/hi/science/nature/4449711.stm Law that has driven digital life: The Impact of Moore's Law] – A comprehensive BBC News article, 18 April 2005
| |
| * [http://www-03.ibm.com/press/us/en/pressrelease/19260.wss IBM Research Demonstrates Path for Extending Current Chip-Making Technique] – Press release from IBM on new technique for creating line patterns, 20 February 2006
| |
| | |
| ===Data===
| |
| *[http://www.jetrepo.com/compsci/cpu_speed.pdf Intel (IA-32) CPU Speeds] since 1994. Increases in recent years have seemed to slow down (in terms of percentage increase per year).
| |
| | |
| ===FAQs===
| |
| * [http://news.com.com/FAQ+Forty+years+of+Moores+Law/2100-1006_3-5647824.html?tag=nefd.lede a C|net FAQ about Moore's Law]
| |
| | |
| [[Category:CZ Live]] | |
Moore's law is the prediction that the transistor density inside integrated circuits will double every two years, and that prices will decline at the same time[1].
The phenomemon predicted by Moore's law, first described in 1965, has held remarkably true to date, and experts predict that this trend might continue until ~2020 or so, declining at the point where switching element sizes reach the molecular level. Moore's law is not really a law, but rather more a “rule of thumb” or a practical way to think about something.
Another way to think about it is that density is inversely proportional to the distance that signals must traverse, so Moore's Law does not only address the computational capability of more circuit elements, but workarounds to speed-of-light limitations.
186,300 miles per second. It's not just a good idea. It's the Law.
— Seen at a T-shirt at an IETF meeting
Moore's law is named for Gordon Moore, a co-founder of Intel, who wrote about it in "Cramming more components onto integrated circuits", Electronics Magazine 19 April 1965[1]:
The complexity for minimum component costs has increased at a rate of roughly a factor of two per year ... Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years. That means by 1975, the number of components per integrated circuit for minimum cost will be 65,000. I believe that such a large circuit can be built on a single wafer
Although named for him, Gordon Moore may not have invented Moore's law; instead, he may have heard Douglas Engelbart, a co-inventor of the mechanical computer mouse, discuss the projected downscaling of integrated circuit size in a 1960 lecture.[2]. Moore's observation was named a 'law' by the Caltech professor and VLSI pioneer Carver Mead[1].
In 1975, Moore projected a doubling only every two years. He is adamant that he never said "every 18 months", but that is how it has been quoted. The SEMATECH roadmap follows a 24 month cycle. In April 2005, Intel offered $10,000 for a copy of the original Electronics Magazine. [3]
It has become common practice to cite Moore's Law as a predictor for the rapid advance in computing power per unit cost in a variety of computer-related technologies, such as hard disk storage cost per unit of information, even though such advances may have little to do with transistor technology advances. Moore's Law has become synonymous with saying "things very quickly gain capabilities, while cost falls".
Processor speed not always the limiting factor
A new important chapter started for Moore's law in 2003. The cycles per second for microprocessors went flat, and processor designers began to keep Moore's law alive by adding
extra cores. multicore microprocessors in personal computers. [4]
Power consumption
The problem of absolute power consumption surfaced at about the same time.
The power consumption of computers is now a non-trivial fraction of each modern nation's total power output. (about 2% in 2005) The current limitations to Moore's law are not due to hardware engineering, rather the burden is on programmers to make use of multiple cores and the burden is on society to pay the cost of the extra electrical load. In response, metrics like flops/watt are being used to help keep computation affordable. Note: 1.2% of total US electrical output in 2005 went for servers alone. http://dl.klima2008.net/ccsl/koomey_long.pdf (Koomey, 2008)
Processor faster; software slower
There is a joke about "Gates' Law", a sort of inverse of Moore's Law:
“The speed of software halves every 18 months.” This oft-cited law is an ironic comment on the tendency of software bloat to outpace the every-18-month doubling in hardware capacity per dollar predicted by Moore's Law. The reference is to Bill Gates; Microsoft is widely considered among the worst if not the worst of the perpetrators of bloat. Jargon file
Memory
Some applications are memory, not processor limited. Admittedly, memory is also dependent on many of the same semiconductor technologies as processors, but a given computer may not be able to accept more physical memory. Desktop computers with Microsoft operating systems often improve more with more memory than a replacement faster processor -- although the situation can be confusing if a different processor is needed to accept more memory.
In the early 1990s, the crisis in Internet routing, which was worked around with Classless Inter-Domain Routing (CIDR), was a practical consequence of the Cisco AGS, then the most common Internet core router, holding a maximum of 16 megabytes of memory. Until the CIDR techniques went into operational use, the number of routes was doubling every five months and threatening to overload the routing tables of those routers.
Networks
Another set of laws deals with the growth of connectivity in computer networks. Sarnoff's Law deals with the value of a one-to-many radio or television broadcast network, as opposed to Metcalfe's Law about any-to-any internets.
References
