Brief review on future scenario of Processor Design Technologies

Chinmay sanjay chepurwar Alhad peshwe

Address
1

[email protected]

Contact no-8087729996
Abstract The present paper intends to review the future scenario of miniature sized high performance processors involving different design technologies. Nano devices and nano gadgets are need of an hour since the size of devices expected from industry are getting smaller and smaller with highly improved performance. The upcoming cutting edge technologies will require more computing power with higher capabilities and reduced size. To fulfil these demands various processor design technologies play a vital role. CMOS VLSI technology is a revolution in minimizing the size of processors. But for further extension of Moores law Intel Corporation made a major breakthrough in VLSI technology in the form of TRI-GATE transistors in MAY 2011. TRI-GATE transistors technology will replace the size of present 32 nM transistor to 22nM. But this will extent Moores law only to few years and size to 14 nM. To further extend Moores law and to reduce size, Graphene transistors will play key role. Further increment in number of transistors will begin to hinder their performance. This will surely lead us to enter into a new era of Quantum computing. These three design technologies are discussed in this paper.

Keywords TRI-GATE transistors, graphene, Quantum computing

I. INTRODUCTION Transistors are the microscopic, silicon-based switches that process the ones and zeros of the digital worlds and are the fundamental BUILDING BLOCK of all semiconductor chips. With traditional planar transistors, ELECTRONIC signals travel as if on a flat, one-way road. This approach has served the semiconductor industry well since the 1960S . In 1965 Gordan Moore predicted that number of transistors on a chip will double roughly every year(a decade later, revised to every two years) in the future.This law was followed by all processor design companies which led to the shrinking of transistors upto 33nM in i7 core processors. But after i7 processors, the density of transistors was so high that with current technologies it was not possible to

shrink the transistors further. Hence to continue Moores law Intel launched its TRI GATE transistors technology in May 2011. TRI-GATE or 3-D are terms used by Intel Corporation to describe the nonplanar transistor architecture planned for use in future microprocessors. These transistors employ a single gate stacked on top of two vertical gates allowing for essentially three times the surface area for electrons to travel. Intel reports that their tri-gate transistors reduce leakage and consume far less power than current transistors. This allows up to 37% higher speed, and power consumption at under 50% of the previous type of transistors used by Intel. The tri-gate is built on an ultra-thin layer of fully depleted silicon for reduced current leakage. This allows the transistor to turn on and off faster, while dramatically reducing power consumption. It also incorporates a raised source and drain structure for low resistance, which allows the transistor to be driven with less power. The design is also compatible with the future introduction of a high K gate dielectric for even lower leakage. This technology can be used to shrink the transistor size to about 14nM only. For further shrinkage it would be necessary to search for a better semiconductor material for designing of processors. According to researchers Graphene is considered to be the best option in sight. Graphene, a single atomic layer of carbon which is 2D crystalline hexagonal in structure has a number of unique properties. top of each other, but it was believed that these layers of carbon cannot be separated since they will become unstable and it was believed that they would never get its isolated form. But in 2004 Konstantin Novoselov ,Andre Geim and their collaborator showed that single layer stable carbon layer could be isolated . This layer is called as graphene. Graphene is a single layer of carbon packed in hexagonal 2D lattice with a carbon carbon atomic distance of 0.142 nM. This size and some of its unique properties make Graphene the best candidate for

replacing silicon. This will surely extend principle of Gordan Moore to some extent. Gershenfeld says that if making transistors smaller and smaller is continued with the same rate as in the past years, then by the year of 2020, the width of a wire in a computer chip will be no more than a size of a single atom. These are sizes for which rules of classical physics no longer apply and laws of quantum mechanics will play a key role in designing processors . According to study, around 2030 computers might not have any transistors and chips since depth operation of processors would reach subatomic particles . Computers designed on today's chip technology will not continue to get cheaper and better. Because of its great power, QUANTUM COMPUTER is an attractive next step in computer technology. Unit of information will no longer be a bit, but a qubit. A technology of quantum computers is also very different. For operation, quantum computer uses quantum bits (qubits). Qubit has a quaternary nature. Quantum mechanics laws are completely different from the laws of a classical physics. A qubit can exist not only in the states corresponding to the logical values 0 or 1 as in the case of a classical bit, but also in a superposition state.

May 2011: Intel announces the introduction of TRI GATE transistors.

III TRI GATE TRANSISTORS IN DETAIL.

II HISTORY
Since the invention of transistors in 1947 technology has progressed swiftly, paving the way for ever more powerful yet more cost effective and energy efficient products. Continuation of these advances at the pace dictated by Moores law has required numerous innovations. Recent notable ones are strained silicon or high k/ metal gates. Intel is now about to make yet another radical change in transistor design, one that will deliver an unprecedented and combination of performance and energy efficiency. Lets take a look at transistor history and key milestones.

Dec 16 ,1947: William Shockley, John Bardeen successfully built the first transistor at Bell Labs 1950; William Schokley developed a BJT. Oct 18,1954: First transistor Radio was put on market April 25, 1961: First patent was awarded to Robert Noyce for ICs. 1965: Moores law 1969: Intel develops first successful P MOS silicon gate transistor technology 1971: Intel launched its first microprocessor 4004 with 10 microns P MOS technology 1985: Intel 386 microprocessor was released with 275000 transistors using 1.5 microns C MOS technology Aug13, 2002: Intel unveils 90nM process technology Jan 2007: Intel introduced high k and metal gate transistors with size of 45nM.

Intel's tri-gate transistor employs a novel 3-D structure, like a raised, flat plateau with vertical sides, which allows electronic signals to be sent along the top of the transistor and along both vertical sidewalls as well. This effectively triples the area available for electrical signals to travel, like turning a one-lane road into a three-lane highway, but without taking up more space. Besides operating more efficiently at nanometer-sized geometries, the tri-gate transistor runs faster, delivering 20 percent more drive current than a planar design of comparable gatesize. Intel researchers have developed "tri-gate" transistor design. This is one of the major breakthroughs in the VLSI technology. The transistor is aimed at bringing down the transistor size in accordance with the Moores Law. The various problems transistors with very small size face have to be overcome. A reduction in power dissipation is another aim. This is to develop low power micro processors and flash memories. The tri-gate structure is a promising approach for extending the Terahertz transistor architecture Intel announced in December 2001. The tri-gate is built on an ultra-thin layer of fully depleted silicon for reduced current leakage. This allows the transistor to turn on and off faster, while dramatically reducing power consumption. It also incorporates a raised source and drain structure for low resistance, which allows the transistor to be driven with less power. The design is also compatible with the future introduction of a high K gate dielectric for even lower leakage. Tri-gate transistors show excellent DIBL, high sub threshold slope, high drive and much better short channel performance compared to CMOS bulk transistor. The drive current is almost increased by 30%. The thickness requirement of the Si layer is also relaxed by about 2-3 times that of a CMOS bulk transistor. Tri- gate transistors are expected to replace the nanometre transistors in the Intel microprocessors by 2010. 60 nm tri-gate transistors are already fabricated and 40 nm tri-gate transistors are under fabrication. Trigate transistor is going to play an important role in decreasing the power requirements of the future processors. It will also help to increase the battery life of the mobile devices.

Tri-Gate transistors down to 22nm were fabricated in the following manner. To get body widths of the same approx. size as the polysilicon gate, the body was first fabricated by treating it in a similar manner to

polysilicon, using aggressive poly-silicon lithography and etch techniques to get body thicknesses equal to gate lengths. The body was then doped to obtain acceptable threshold voltages (Vt) using conventional boron implants.

No halo implants were used for setting Vt, nor were there any angled implants used anywhere in the process. This is in contrast to Double-Gate (DG), and this is possible since the Tri-Gate very much resembles bulk transistor from the processing point-of-view. However, to get the right Vts, it was found necessary to protect the Tri-Gate bodies from boron out diffusion into the surrounding oxide by an N2O oxidation before gate definition. The gate stack included polysilicon gates, and a conventional physical oxide thickness of 15 Angstroms. Raised source/drains were used to reduce parasitic resistances and the transistor was silicided using nickel.CMOS Tri-Gate transistors were fabricated down to 30nm.The examples of CMOS devices at Lg=60nm. the cross-section of the nMOS device . This device has body dimensions of HSi=36nm and WSi=55nm, The NMOS device had a subthreshold slope (S/S) = 68 mV/decade, DIBL=41mV/V, Ion=1.14mA/mm and Ioff=70nA/mm at Vcc= 1.3V. The PMOS device showed S/S=69.5 mV/decade, DIBL= 48mV/V, Ion=520mA/mm and Ioff = 24nA/mm at Vcc=1.3V. The tri-gate is built on an ultra-thin layer of fully depleted silicon for reduced current leakage. This allows the transistor to turn on and off faster, while dramatically reducing power consumption. It also incorporates a raised source and drain structure for low resistance, which allows the transistor to be driven with less power. The design is also compatible with the future introduction of a high K gate dielectric for even lowerleakage. Moore's law describes a long-term trend in the history of computing hardware, in which the number of transistors that can be placed inexpensively on an integrated circuit has doubled approximately every two years.Moore's law 21st centuries precisely describes a driving force of technological and social change in the late 20th and early.

IV GRAPHENE IN DETAIL

Carbon is arguably THE MOST fascinating ELEMENT IN THE PERIODIC table.. the most common form of carbon is graphite, which consists of stacked sheets of carbon with hexagonal structure. a new form OF MOLECULAR carbons IS the so called fullerenes. the most common, called c60, contains 60 carbon atoms and looks like a FOOTBALL (soccer ball) made UP

FROM 20 HEXAGONS AND 12 PENTAGONS WHICH ALLOW THE SURFACE TO FORM A SPHERE.. A RELATED QUASSIONE DIMENSIONAL FORM OF CARBON,CARBON NANOTUBES, HAVE BEEN KNOWN FOR SEVERAL DECADES AND THE SINGLE WALLED NANOTUBES SINCE 1993.THESE CAN BE FORMED FROM GRAPHENE SHEETS WHICH ARE ROLLED UP TO FORM TUBES, AND THEIR ENDS ARE HALF SPHERICAL IN THE SAME WAY AS THE FULLRENES. THE ELECTRONIC AND MECHANICAL PROPERTIES OF METALLIC SINGLE WALLED NANOTUBES HAVE MANY SIMILARITIES WITH GRAPHENE. It was well known that graphite consists hexagonal sheets staked on the top of each other, but it was believed that these layers of carbon cannot be separated since they will become unstable and it was believed that they would never get its isolated form. But in 2004 konstantin novoselov ,adregeim and their collaborator showed that single layer stable carbon layer could be isolated . This layer is called as graphene. Graphene is a single layer of carbon packed in hexagonal 2D lattice with a carbon carbon atomic distance of 0.142 nM. It is first truly two dimensional crystalline materials it is representative of a whole 2D materials. Several other 2D crystals were made with the help of this crystal. The electronic structure of graphene is different from usual three dimensional. Its fermi surface is characterised by six double cones. In intrinsic graphene conductivity is quite low due to density of state of material is zero at cone points. The fermi level can be changed by applying electric field thus it become n doped or p doped graphene depending on polarity of field. It can be doped by adsorbing on its surface. The electrical conductivity of doped graphene is potentially quite high; at room temperature it is far greater than copper. Graphene is practically transparent. In the optical region it absorbs only 2.3% of light. It is stronger than steel and stretchable than any other material. The unusual properties and size make graphene a excellent candidate as a basic material in electronics to replace semiconductors. It also removes the several limitations of semiconductors. Due to small size and good conductivity it will definitely play key role in reducing the size of electronic gadgets. It will definitely change the face of electronics in real aspects. The recent development in this field is production of transistors and diodes. Even IBM has announced that they will produce graphene transistor very soon on large scale.

V A QUICK VIEW OF TRI-GATE TRANSISTOR AND GRAPHENE

GRAPHENE Source google

32nM planar transistor

ChurchTuring thesis. However, in practice infinite resources are never available and the computational basis of 500 qubits, for example, would already be too large to be represented on a classical computer because it would require 2500 complex values to be stored. .Nielsen and Chuang point out that "Trying to store all these complex numbers would not be possible on any conceivable classical computer."

VII ADAVANTAGES
TRI-GATE TRANSISTORS:

22nM TRI GATE Transistors

Lower leakage and consume much less power. Faster & cooler operation. 45% increase in speed or 50x reduction in off-current. Moore's Law scaling can be taken well into the next decade . The basic building blocks for future microprocessors Process cost adder 2-3% 37% performance increase at low voltage >50% power reduction at constant performance. Same transistor and interconnect features as on 22 nm CPUs . GRAPHENE: Low cost Excellent thermal stability More compact size Higher speed Higher mechanical strength. Less leakage current. Very High switching speed.

VI IDEA OF QUANTUM COMPUTERS A quantum computer is a device for computation that makes direct use of quantum mechanical phenomena, such as superposition and entanglement, to perform operations on data. Quantum computers are different from traditional computers based on transistors. The basic principle behind quantum computation is that quantum properties can be used to represent data and perform operations on these data. A theoretical model is the quantum Turing machine, also known as the universal quantum computer. Quantum computers share theoretical similarities with non-deterministic and probabilistic computers, like the ability to be in more than one state simultaneously. The field of quantum computing was first introduced by Richard Feynman in 1982. Although quantum computing is still in its infancy, experiments have been carried out in which quantum computational operations were executed on a very small number of qubits (quantum bits). Both practical and theoretical research continues, and many national government and military funding agencies support quantum computing research to develop quantum computers for both civilian and national security purposes, such as cryptanalysis.[3] Large-scale quantum computers could be able to solve certain problems much faster than any classical computer by using the best currently known algorithms, like integer factorization using Shor's algorithm or the simulation of quantum manybody systems. There exist quantum algorithms, such as Simon's algorithm, which run faster than any possible probabilistic classical algorithm. Given unlimited resources, a classical computer can simulate an arbitrary quantum algorithm so quantum computation does not violate the

Quantum Processor: Ultra high speed Very High switching Parallelism with single processor Improved networking Ability to perform some algorithms which classical computers cannot perform.

VIII CONCLUSION
The present paper intended to focus on the future prospects and different technologies associated with processor designing .TRI-GATE transistor technology will surely extend Moores law for a decade which will then be followed by Graphene transistors. After the era of Graphene transistors the next generation computers will be powered by quantum computers. With todays technology

quantum computers seems to be a distant dream hence lot of research and development is needed in this field. Fabrication of graphene ICs should be promoted and new techniques should be introduced for better performance of Graphene processors.

REFERENCES
[1] Geim AK and novoselov, KS (2007).the rise of graphene- nature materials 6. [2]mechanical and electrical properties of graphene by joseph scott bunch(cornell university) [3] wikipedia [4] www.intel.com [5]images-www.google.com. [6] embedded for you- December 2011