MSE-630 Integrated Circuits and the Future of Semiconductors.

Slides:



Advertisements
Similar presentations
EXPLORING QUANTUM DOTS
Advertisements

DRAFT - NOT FOR PUBLICATION 14 July 2004 – ITRS Summer Conference ITRS FEP Challenges Continued scaling will require the introduction of new materials.
by Alexander Glavtchev
An International Technology Roadmap for Semiconductors
Single Electron Devices Single-electron Transistors
Metal Oxide Semiconductor Field Effect Transistors
1 Cleared for Open Publication July 30, S-2144 P148/MAPLD 2004 Rea MAPLD 148:"Is Scaling the Correct Approach for Radiation Hardened Conversions.
Logic Process Development at Intel
ECE 6466 “IC Engineering” Dr. Wanda Wosik
CHALLENGES IN EMBEDDED MEMORY DESIGN AND TEST History and Trends In Embedded System Memory.
Introduction to CMOS VLSI Design Lecture 21: Scaling and Economics
Lecture 2: Modern Trends 1. 2 Microprocessor Performance Only 7% improvement in memory performance every year! 50% improvement in microprocessor performance.
Integrated Circuits (ICs)
Integrated Digital Electronics Module 3B2 Lectures 1-8 Engineering Tripos Part IIA David Holburn January 2006.
11/29/2004EE 42 fall 2004 lecture 371 Lecture #37: Memory Last lecture: –Transmission line equations –Reflections and termination –High frequency measurements.
Nanoscale structures in Integrated Circuits By Edward Mulimba.
Design and Implementation of VLSI Systems (EN0160)
S. Reda EN160 SP’08 Design and Implementation of VLSI Systems (EN1600) Lecture 18: Scaling Theory Prof. Sherief Reda Division of Engineering, Brown University.
Institute of Digital and Computer Systems 1 Fabio Garzia / Finding Peak Performance in a Process23/06/2015 Chapter 5 Finding Peak Performance in a Process.
For the exclusive use of adopters of the book Introduction to Microelectronic Fabrication, Second Edition by Richard C. Jaeger. ISBN © 2002.
How engineers in 1954 expected a computer to look like in 2004.
3.1Introduction to CPU Central processing unit etched on silicon chip called microprocessor Contain tens of millions of tiny transistors Key components:
MonolithIC 3D Inc., Patents Pending MonolithIC 3D ICs RCAT approach 1 MonolithIC 3D Inc., Patents Pending.
{ The demise of Conventional Computing? Bilal Kaleem, 1 st Year Physics.
General Licensing Class Oscillators & Components Your organization and dates here.
G.K.BHARAD INSTITUTE OF ENGINEERING DIVISION :D (C.E.) Roll Number :67 SUBJECT :PHYSICS SUBJECT CODE : Presentation By: Kartavya Parmar.
Device Physics – Transistor Integrated Circuit
Keeping Up with Moore’s Law Who the heck is this Moore guy anyway? Gordon E. Moore was the cofounder of Intel Corporation Gordon E. Moore was the cofounder.
Lecture 03: Fundamentals of Computer Design - Trends and Performance Kai Bu
CMOS Scaling Two motivations to scale down
History of Integrated Circuits  In 1961 the first commercially available integrated circuits came from the Fairchild Semiconductor Corporation.  The.
Carrier Mobility and Velocity Mobility - the ease at which a carrier (electron or hole) moves in a semiconductor Mobility - the ease at which a carrier.
Multi-core Programming Introduction Topics. Topics General Ideas Moore’s Law Amdahl's Law Processes and Threads Concurrency vs. Parallelism.
C OMPUTER O RGANIZATION AND D ESIGN The Hardware/Software Interface 5 th Edition Chapter 1 Computer Abstractions and Technology Sections 1.5 – 1.11.
Limitations of Digital Computation William Trapanese Richard Wong.
Text Book: Silicon VLSI Technology Fundamentals, Practice and Modeling Authors: J. D. Plummer, M. D. Deal, and P. B. Griffin Class: ECE 6466 “IC Engineering”
Computer Evolution. ENIAC - background Electronic Numerical Integrator And Computer Eckert and Mauchly University of Pennsylvania Trajectory tables for.
Introduction to ICs and Transistor Fundamentals Brief History Transistor Types Moore’s Law Design vs Fabrication.
W E L C O M E. T R I G A T E T R A N S I S T O R.
Semiconductor Industry Milestones
FPGA-Based System Design: Chapter 1 Copyright  2004 Prentice Hall PTR Moore’s Law n Gordon Moore: co-founder of Intel. n Predicted that number of transistors.
EE141 © Digital Integrated Circuits 2nd Introduction 1 Principle of CMOS VLSI Design Introduction Adapted from Digital Integrated, Copyright 2003 Prentice.
Trends in IC technology and design J. Christiansen CERN - EP/MIC
Moore’s Law and Its Future Mark Clements. 15/02/2007EADS 2 This Week – Moore’s Law History of Transistors and circuits The Integrated circuit manufacturing.
Nanometer Technology © Copyright 2002, Fairview Ridge Partners, LLC All Rights Reserved Nanometer Technology AKI Expert Session.
Caltech CS184 Winter DeHon 1 CS184a: Computer Architecture (Structure and Organization) Day 6: January 22, 2003 VLSI Scaling.
Overview of VLSI 魏凱城 彰化師範大學資工系. VLSI  Very-Large-Scale Integration Today’s complex VLSI chips  The number of transistors has exceeded 120 million 
DR. SIMING LIU SPRING 2016 COMPUTER SCIENCE AND ENGINEERING UNIVERSITY OF NEVADA, RENO Session 3 Computer Evolution.
The Fate of Silicon Technology: Silicon Transistors Maria Bucukovska Scott Crawford Everett Comfort.
CS203 – Advanced Computer Architecture
William Stallings Computer Organization and Architecture 6th Edition
CS203 – Advanced Computer Architecture
Introduction to VLSI ASIC Design and Technology
Technology advancement in computer architecture
Morgan Kaufmann Publishers
Architecture & Organization 1
EE 4611 INTRODUCTION 21 January 2015 Semiconductor Industry Milestones
IC TECHNOLOGY.
Architecture & Organization 1
Overview of VLSI 魏凱城 彰化師範大學資工系.
Device Physics – Transistor Integrated Circuit
3.1 Introduction to CPU Central processing unit etched on silicon chip called microprocessor Contain tens of millions of tiny transistors Key components:
PRESENTATION ON TRI GATE TRANSISTOR PREPARED BY: SANDEEP ( )
Device Physics – Transistor Integrated Circuit
Semiconductor Industry:
Reading (Rabaey et al.): Sections 3.5, 5.6
Technology scaling Currently, technology scaling has a threefold objective: Reduce the gate delay by 30% (43% increase in frequency) Double the transistor.
Unit -4 Introduction to Embedded Systems Tuesday.
Presentation transcript:

MSE-630 Integrated Circuits and the Future of Semiconductors

MSE-630 The Electronics Industry is BIG Business!! In 2008, global sales of semiconductor devices and components are projected to be $309 billion When incorporated into end-use components built on semiconductor devices, the market is projected to be 1.7 trillion dollars

MSE-630 The Revenues of U.S.-based chip companies account for nearly half of global semiconductor sales and more than three-quarters of U.S.-owned chip manufacturing Capacity is located in the United States. The U.S. chip industry provides more than $100 mission annually to support research in U.S. Universities, invests $15 billion in R&D, and employs 226,000 people Technology exports account for 23% of total exports. 75% of the chip industry revenue is from export sales.

MSE-630 Historically, as the speed of devices increases, their size decreases The transistors manufactured today are 20 times faster and occupy less than 1% of the area of those built 20 years ago In a 1965 paper, Gordon Moore stated that the number of components on the most complex integrated circuit chip would double each year for the next ten years In 1965, there were components on the average chip

MSE nm Logic Technology Intel recently disclosed details of its 65 nm generation logic technology which includes numerous features to improve performance and reduce power. This technology is being demonstrated on fully functional 70 Mbit SRAM chips with over 1/2 billion transistors. Once again proving that Moore’s Law is alive and well, Intel’s 65 nm technology is on track for delivery in 2005.recently disclosed

MSE-630 To continue to grow, the semiconductor industry has to overcome several technological challenges: Lithography Transistor scaling Interconnections Circuit families Computer memory Circuit design

MSE-630 Moore pointed out in his original paper that the doubling 0f the number of components on an integrtated circuit was due to three factors: Half the increase is derived from improvement in lithographic resolution A quarter comes from larger chip sizes, made possible by improved manufacturing techniques and getter lithography The remaining 25% is due to innovation, such as more creative techniques for forming components on a chip The industry will continue to grow so long as the rate of increase of components and functions on a chip exceeds the rate of increase of the cost per chip.

MSE-630 Lithography Historical and future trends of lithographic resolution capability Originally mercury lamps were used to get resolution of ~350 nm Deep ultraviolet reduces dimensions to 250-nm. Smaller wavelengths present challenges due to diffraction of light and distortion, as well as photoresist materials Electron-beam lithography can, with = 0.01-nm, has high resolution. Key problems include: -multiplicity of masks required -mask integrity -cost

MSE-630 Lithography Proximity X-ray lithography has been used to fabricate ICs to 150-nm. The primary problem is that lenses and mirrors are not available for these wavelengths. Blocking masks must be used with features of the same dimension as that on the wafer. The cost and difficulty of fabricating thee masks without distortion are key challenges The biggest risk of any new lithographic technique is that the benefits derived from increased component density are outweighed by the increased cost.

MSE-630 Transistor Scaling and Design The key tradeoff in technology adaptation is speed vs. cost. Bipolar transistors are faster than CMOS, but CMOS has higher circuit density – thus, CMOS wins out Comparison of projected vs. actual device performance Trend of microprocessor clock frequency For any reduction  in linear dimensions the voltage and doping levels can be adjusted to increase performance by  and decrease power density by  2

MSE-630 Transistor Scaling and Design Because devices operate at room temperature, off current limits designs to threshold voltages of 0.3V or higher Future improvements will require significantly lower operating temperatures Continued speed increases will require improved software and I/O design

MSE-630 Plausible evolution in transistor structure toward a more symmetric structure that results in better control of the fields in the gate region, regulating device condition. The FETs pictured are: (a) bulk Si, (b) silicon-0n-insulator (SOI), (c) ground plane, counter electrode (d) verticle double gate and (e) fully symmetric double gate Shrinking components results in gate-oxide tunneling. The limit of a useful device with an on/0ff current ratio of 1000 due to source-drain tunneling alone appears to be about 5-nm separation between source and drain. Accounting for dopant fluctuations, the lower limit is likely 10-nm. Possible advances in can come from: Shorter channel lengths Materials with higher performance

MSE-630 New device designs will move toward three- dimensional arrays of devices. These devices reduce space while using the same size components as current devices Transistor Scaling and Design

MSE-630 Low Temperature Operation Operating at LN2 temperatures (77K) would improve performance by a factor of 2. Problems include: Refrigerator cost Reliability Need to redesign technology to optimize low-temperature operation Optimum cost/performance operation may occur at -50 o C using thermoelectric methods The best candidates for this are high-end servers

MSE-630 Wiring and Connections Traditional interconnects are Al or Cu – both of which have resistance and capacitance Using materials with low dielectric constants in insulation layers allows continued decreases in size Another potential solution is to use a hierarchical wiring scheme, which combines high-density wiring at the first few levels with larger, lower-resistance and capacitance wires at upper levels

MSE-630 Circuit Families Bipolar vs CMOS performance trends Bipolar transistors win on speed – CMOS wins on device density The large number of components on a chip lead to a superior system performance and lower cost per components. With more devices, more functions can be designed into a single chip. Bipolar transistors generate more heat than CMOS

MSE-630 No alternative logic technology is evolving within cost/performance products on the market today to threaten CMOS dominance. In light of the continuing CMOS performance evolution, an even steeper evolution and learning curve would be required to displace CMOS. No radical shift in circuit type seems to be on the horizon. Moore’s law will continue to hold for approximately 10 more years

MSE-630 Memory Cells DRAM is a cell consisting of a transistor and capacitor. For the past 20 years, DRAM products have followed a generational evolution leading to a 4X increase in bits per chip every three years. The current limit in size is 4X the square of the lithographic dimension As lithographic improvements slow, so will growth of Memory In 1990, 1Mb of memory cost ~$175 retail. Now, a 1Gb DIMM costs $130

MSE-630 Design The increase in function – the ways the devices on a chip can be arranged, will be the driving force for new ICs, not the sheer number of devices on a chip For a 10GHz processor, the clock cycle time is 100 ps. Since light travels at 300  m/ps, in vacuum, the space reachable by light in one clock cycle is 30-mm. Assuming a medium consisting of typical dielectrics rather than vacuum, the reachable space is of the order of mm, roughly the size of today’s chips. This places an upper bound on clock speeds and planar chip sizes

MSE-630 Cost Cost reduction is a major tenet of Moore’s law. The primary factor underlying the decreasing cost per circuit or memory bit is the increase in density, or circuits per square millimeter. The cost of processing a silicon wafer must increase much less rapidly than the density in order to achieve cost reduction. The rate of cost increase in the silicon chip manufacturing is approximately 15% per year. This si sdue to: Stabilization of clean room requirements Better equipment productivity and utilization Slower increase in the number of process steps The rate of increase in costs must be matched by a greater rate of increase of components per chip to continue to thrive

MSE-630 Future Directions CMOS technology is likely to continue to evolve and dominate the semiconductor industry for the next years Optical lithography must be extended to unanticipated levels or be replaced by non-optical techniques. Transistors must be replaced with a radical new structure using new materials DRAM cells must be designed in as-yet-unknown structures to achieve economically viable increase in memory chip integration Wires must be fabricated at tenth-of-a-micron dimensions in hierarchical structure with low-dielectric constant materials Dynamic circuits and SRAM cells must be designed to provide more function for a given set of transistors Cost reductions will continue to be driven by the ability to integrate more functions on a chip

MSE-630