Low Energy Communication: NanoPhotonic & Electrical. Prof. Eli Yablonovitch EECS Dept. UC Berkeley

Similar documents
Lecture 4 INTEGRATED PHOTONICS


Semiconductors, ICs and Digital Fundamentals

Intel's 65 nm Logic Technology Demonstrated on 0.57 µm 2 SRAM Cells

Lecture: Integration of silicon photonics with electronics. Prepared by Jean-Marc FEDELI CEA-LETI

Electronic-Photonic ICs for Low Cost and Scalable Datacenter Solutions

Fundamentals of CMOS Image Sensors

EE70 - Intro. Electronics

Detectors for Optical Communications

SRM INSTITUTE OF SCIENCE AND TECHNOLOGY (DEEMED UNIVERSITY)

Figure Responsivity (A/W) Figure E E-09.

ATV 2011: Computer Engineering

Downloaded from

Lecture #29. Moore s Law

Figure Figure E E-09. Dark Current (A) 1.

Department of Electrical Engineering IIT Madras

Electronic Components (Elements)

Optical Receivers Theory and Operation

NEXT GENERATION SILICON PHOTONICS FOR COMPUTING AND COMMUNICATION PHILIPPE ABSIL

Nanotechnology, the infrastructure, and IBM s research projects

The Past, Present, and Future of Silicon Photonics

AE103 ELECTRONIC DEVICES & CIRCUITS DEC 2014

6.012 Microelectronic Devices and Circuits

Chapter 3: Basics Semiconductor Devices and Processing 2006/9/27 1. Topics

Integrated diodes. The forward voltage drop only slightly depends on the forward current. ELEKTRONIKOS ĮTAISAI

Nanophotonics for low latency optical integrated circuits

NON-AMPLIFIED PHOTODETECTOR USER S GUIDE

Optical Communications

Chapter Semiconductor Electronics

Semiconductor Devices

Chapter 3 Basics Semiconductor Devices and Processing

Reg. No. : Question Paper Code : B.E./B.Tech. DEGREE EXAMINATION, NOVEMBER/DECEMBER Second Semester

Silicon Photonics Technology Platform To Advance The Development Of Optical Interconnects

Electronic Components

Transistor Characteristics

Silicon Photonics Photo-Detector Announcement. Mario Paniccia Intel Fellow Director, Photonics Technology Lab

Design of Nano-Electro Mechanical (NEM) Relay Based Nano Transistor for Power Efficient VLSI Circuits

MOSFET short channel effects


14.2 Photodiodes 411

NON-AMPLIFIED HIGH SPEED PHOTODETECTOR USER S GUIDE

Physics 160 Lecture 5. R. Johnson April 13, 2015

Lecture 1 Introduction to Solid State Electronics

10/27/2009 Reading: Chapter 10 of Hambley Basic Device Physics Handout (optional)

4/30/2012. General Class Element 3 Course Presentation. Circuit CoCircuit Componentsmponents. Subelement G6. 3 Exam Questions, 3 Groups

Unit 2 Semiconductor Devices. Lecture_2.5 Opto-Electronic Devices

ECE 5745 Complex Digital ASIC Design Topic 2: CMOS Devices

Diodes Rectifiers, Zener diodes light emitting diodes, laser diodes photodiodes, optocouplers

Lecture 2. Part 2 (Semiconductor detectors =sensors + electronics) Segmented detectors with pn-junction. Strip/pixel detectors

A silicon avalanche photodetector fabricated with standard CMOS technology with over 1 THz gain-bandwidth product

UNIT-VI FIELD EFFECT TRANSISTOR. 1. Explain about the Field Effect Transistor and also mention types of FET s.

for optical communication system

Electromagnetic spectrum

High speed silicon-based optoelectronic devices Delphine Marris-Morini Institut d Electronique Fondamentale, Université Paris Sud

Class XII - Physics Semiconductor Electronics. Chapter-wise Problems

MICROPROCESSOR TECHNOLOGY

Ultra-sensitive SiGe Bipolar Phototransistors for Optical Interconnects

Where detectors are used in science & technology

INTERNATIONAL JOURNAL OF APPLIED ENGINEERING RESEARCH, DINDIGUL Volume 1, No 3, 2010

Quantum Condensed Matter Physics Lecture 16

Resonant Tunneling Device. Kalpesh Raval

Silicon Photonics: an Industrial Perspective

1. What is the unit of electromotive force? (a) volt (b) ampere (c) watt (d) ohm. 2. The resonant frequency of a tuned (LRC) circuit is given by

UNIT IX ELECTRONIC DEVICES

UNIT VIII-SPECIAL PURPOSE ELECTRONIC DEVICES. 1. Explain tunnel Diode operation with the help of energy band diagrams.

EE 5611 Introduction to Microelectronic Technologies Fall Thursday, September 04, 2014 Lecture 02

OBJECTIVE TYPE QUESTIONS

Chap14. Photodiode Detectors

Heinrich-Hertz-Institut Berlin

Chapter 3 OPTICAL SOURCES AND DETECTORS

Beyond Transistor Scaling: New Devices for Ultra Low Energy Information Processing

Transistor-Based Ge/SOI Photodetector for Integrated Silicon Photonics. Xi Luo. A dissertation submitted in partial satisfaction of the

WINTER 14 EXAMINATION. Model Answer. 1) The answers should be examined by key words and not as word-to-word as given in the

Pramoda N V Department of Electronics and Communication Engineering, MCE Hassan Karnataka India

Chapter 1: Semiconductor Diodes

Active Technology for Communication Circuits

A 90 nm High Volume Manufacturing Logic Technology Featuring Novel 45 nm Gate Length Strained Silicon CMOS Transistors

Si CMOS Technical Working Group

MOSFET Parasitic Elements

Metal-Oxide-Silicon (MOS) devices PMOS. n-type

MAHARASHTRA STATE BOARD OF TECHNICAL EDUCATION (Autonomous) (ISO/IEC Certified) Subject Code: Model Answer Page No: 1/


ELECTRONIC DEVICES AND CIRCUITS

KOM2751 Analog Electronics :: Dr. Muharrem Mercimek :: YTU - Control and Automation Dept. 1 6 FIELD-EFFECT TRANSISTORS

Strip Detectors. Principal: Silicon strip detector. Ingrid--MariaGregor,SemiconductorsasParticleDetectors. metallization (Al) p +--strips

Optical Fiber Communication Lecture 11 Detectors

IENGINEERS- CONSULTANTS QUESTION BANK SERIES ELECTRONICS ENGINEERING 1 YEAR UPTU

Chapter 8: Field Effect Transistors

Semiconductor Detector Systems

Coherent Receivers Principles Downconversion

HOW TO CONTINUE COST SCALING. Hans Lebon

Towards a Reconfigurable Nanocomputer Platform

Advanced Digital Design

Enabling Breakthroughs In Technology

Tests of monolithic CMOS SOI pixel detector prototype INTPIX3 MOHAMMED IMRAN AHMED. Supervisors Dr. Henryk Palka (IFJ-PAN) Dr. Marek Idzik(AGH-UST)

Unit III FET and its Applications. 2 Marks Questions and Answers

Lecture Wrap up. December 13, 2005

PSD Characteristics. Position Sensing Detectors

Sub-Threshold Region Behavior of Long Channel MOSFET

Alternatives to standard MOSFETs. What problems are we really trying to solve?

Transcription:

Low Energy Communication: NanoPhotonic & Electrical Prof. Eli Yablonovitch EECS Dept. UC Berkeley

What is the energy cost of reading out your flash memory?

Read the current going through a resistor, in the presence of noise: Required ( i) ( i) voltage 2 2 V = 2q i f...shot Noise 4kT = f...johnson Noise R = ir >> 2kT / q ~ 50mVolts Signal to Noise Ratio i Required power iv With a safety margin: Energy Consumed = > > ~ 2q f 2q 40 i 2q i f f kt per = 2kT q bit i 2q f = 4kT f processed

Units: ~40kT/bit of information 0.016 atto-joules/bit of information 0.16 nano-watts/gbit/second This is about 10 6 times better than we are doing today!

There are many type of memory possible: 1. Flash 2. SRAM 3. Dram 4. Magnetic Spin 5. Nano-Electro-Chemical Cells 6. Nano-Electro-Mechanical NEMS 7. Moletronic 8. Chalcogenide glass (phase change) 9. Carbon Nanotubes Similarly there are many ways to do logic. But there are not many ways to communicate: 1. Microwaves (electrical) 2. Optical 5

What will be the energy cost, per bit processed? 1. Logic energy cost ~40kT per bit processed 2. Storage energy cost ~40kT per bit processed 3. Communications currently >100,000kT per bit processed. 6

What is the energy cost for electrical communication? V V 2 noise 2 noise R = = 4kT 4kT R f f Signal Energy Noise Power per bit = 4kT per bit All information processing costs ~ 40kT per bit. (for good Signal-to-Noise Ratio) Great! So what s the problem?

The natural voltage range for wired communication is rather low: V V V V V 2 noise 2 noise 2 noise 2 noise noise = 4kT R f 1 = 4kT R RC 1 = 4kT C 4kT q = q C = 4kT / q 123 100mVolts V 1 mvolt q { / C 10µ Volts The natural voltage range for a thermally activated switch like transistors is >>kt/q, eg. ~ 40kT/q or about ~1Volt Voltage Matching Crisis at the nano-scale! If you ignore it the penalty will be (1Volt/1mVolt) 2 = 10 6 The wire wants 1000 electrons at 1mVolt each. (to fulfill the signal-to-noise requirement >1eV of energy) The thermally activated device wants at least one electron at ~1Volt.

The other Moore's Law, for energy per bit function 10 8 Critical Dimension 10µm 1µm 100nm 10nm Gates including wires Gates only Technology Gap 10 7 10 6 10 5 10 4 10 3 10 2 10 Energy per Bit function (kt) 1 Transistor Measurements by Robert Chau, Intel 1960 1980 2000 Year 0.1 2020 2040 2060 Shoorideh and Yablonovitch, UCLA 2006

p. 114 "In addition, power is needed primarily to drive the various lines and capacitances associated with the system. As long as a function is confined to a small area on a wafer, the amount of capacitance which must be driven is distinctly limited. In fact, shrinking dimensions on an integrated structure makes it possible to operate the structure at higher speed for the same power per unit area."

l l Repeater Repeater Repeater l l Resistance, R = ρ 2 a Capacitance, C ε r ε o l 2 l RC time RC εrεoρ a RC time = (clock period) /2π 2 2 = ε r ε o ρ l a a l a = clock period 2πε ε ρ r o aspect l ratio of wire a 2 10 4800 12 10 10 sec F/cm 2 10 6 Ωcm

aspect l ratio of wire a 4800 Resistance, R l = ρ a 2 = ρ a l a 4800 2 10 = 45nm = 2000Ω ρ = 4800 a 6 Ωcm Capacitance, C ε r ε o l = ε r ε o a l a C = ε r ε o a 4800 C 7 femto-farads <V 2 > = 4kT R f V signal = 0.56 milli-volts <I 2 > = 4(kT/R) f I signal = 0.25 µamps

Johnson Noise: What about very short wires? V V V V 2 noise 2 noise 2 noise 2 noise = = = = 4kT 4kT R f R 4kT C 4kT q 1 RC q C If 4kT q q C, then the signals could be large enough to be efficiently amplified. If C 2 q kt The Coulomb 10 atto-farads, Blockade Capacitance. For wires less than 1µm, a conventional transistor amplifier configuration may be adequate.

A low-voltage technology, or an impedance matching device, needs to be invented/discovered at the Nano-scale: transistor amplifier with steeper sub-threshold slope hν photo-diode ~1eV nano-transformer - + + + + MEM's switch V G + Cryo-Electronics kt/q~q/c giant magneto-resistance spintronics Cu solid electrolyte Cu Electro-Chemical Switch

in An amplifying transistor as a voltage matching device: Small voltage in Large voltage out out Amplification of weak signals has an energy cost! Amplification of weak signals has a speed penalty! ln{i} Current steeper sub-threshold slope Gate Voltage V g correlated electron motion?

Diffusion current The Zener Diode: E Fv E Fc Bias Voltage Current band to band tunneling

Diffusion current band to band tunneling E F Sharp Step Bias Voltage Current The Esaki Diode:

The Backward Diode as a Switch: The Backward Diode: These have been routinely made in Ge homo-junctions, since the 1960's. E Fv E Fc Sharp Step Current Diffusion current Bias Voltage band to band tunneling

The Backward Diode as a Switch: The sub-threshold slope for tunneling depends on the steepness of the band-edges:

I: Steepness, Low Operating Voltage II: On/Off Ratio >10 5 III: On-State Current Drive Capability, ma/µm or better: ms/µm With tunneling alone, you can pick any two out of the three above, but you can't get all three!

Engineer Bands to Move Together Un-Engineered Engineered

Consider Thermal Vibrations k x k x k z k z k y Uniaxial Strain ε xx k y k x k x k z k z Shear Strain k y ε xy k y Silicon Germanium

Electro-Chemically Driven Metallic Switch: 1nm Cu solid electrolyte Cu - V G +

A low-voltage technology, or an impedance matching device, needs to be invented/discovered at the Nano-scale: transistor amplifier with steeper sub-threshold slope hν photo-diode ~1eV nano-transformer - + + + + MEM's switch V G + Cryo-Electronics kt/q~q/c giant magneto-resistance spintronics Cu solid electrolyte Cu Electro-Chemical Switch

That was Electrical Communication. For longer distances, Optical Communication is needed; How efficient can that be? How many photons are needed? This is mainly determined by the photodetectors.

Two-Dimensional Thin Film Photonic Crystals Cross- Section: Si SiO 2 SiO 2 Si 1µm SOI Si/SiO 2 SOI

Monolithic Integration of Optics and Electronics 8 RF amplifier 1x2 optical switch Optical devices and transistors are constructed side-by-side monolithically in the silicon Luxtera, Inc. Approved for Public Release

CMOS Optical Modulator Performance Rolloff >20GHz 10G 10G Eye Luxtera, Inc. Approved for Public Release

Blazar LUX5010 Multirate 4x10G Optical Active Cable Typical Applications InfiniBand Connectivity SDR/DDR/QDR 10/40 Gigabit Ethernet 2 / 4 / 8 / 10 Gbps Fibre Channel Proprietary Cluster Interconnect Ethernet Local Area Network (LAN) Storage Area Network (SAN) 2.5G / 5G PCI-Express Extension Storage Arrays Optical Backplanes Rack-to-Rack, Shelf-to-Shelf Interconnect Test Equipment

The problem is that optical communications As it is currently practiced uses >10,000 photons/bit of information communicated This is mainly determined by the photodetectors.

Ultra-Low Capacitance, Nano-Photodetectors "the photo-detector without the wire" small size small dark current small size small capacitance small capacitance more voltage/photo-electron, less pre-amp noise eliminate capacitance of connecting wire electrically floating no wire Lowest Possible Capacitance A floating island of Ge acting as a gate on a Si- FET, integrates photodetector and pre-amp in one. For example a 65nm island would have a capacitance ~6.5atto-Farads, a single photo-electron would produce a 25milli-volt gating signal, easily amplified.

Photo-hetero-JFET concept: Source n-si Floating Ge island h + h + e - e - Photons modulate depletion width Drain SiO 2 Silicon This is one of many types of photo-transistors!

Design Principle: Transistor with a Floating Gate p-ge e e e e h h+ h + + h + V F n-si Excess photo-holes attract electrons in the Silicon and change its conductance Open Ge/Si junction gets forward biased by an amount dependant on the light intensity

Reflectors integrated with SOI waveguides to create a transverse cavity n-si channel Ge island Source n-si n-si Drain incident light SiO 2 Silicon

The problem is that optical communications As it is currently practiced uses >10,000 photons/bit of information communicated If the photodetector is small enough, and preamp is sensitive enough, we can anticipate getting this down to: ~15 photons/bit of information communicated For a ~10 3 times improvement

p. 114 "In addition, power is needed primarily to drive the various lines and capacitances associated with the system. As long as a function is confined to a small area on a wafer, the amount of capacitance which must be driven is distinctly limited. In fact, shrinking dimensions on an integrated structure makes it possible to operate the structure at higher speed for the same power per unit area."

Conclusions: 1. Communications is a big bottleneck for reducing powering consumption in information processing. 2. The powering voltage for circuits is close to 1Volt, but it could be reduced to ~1milli-Volt, for a 10 6 reduction in power. 3. Optical communications uses >10,000 photons/bit of information, but with more nanoscopic photodetectors this could be reduced to ~15photons/bit for a 10 3 reduction in power.

Reversible and Adiabatic Computing: Energy/bit function <<kt Charles Bennett, IBM

The other Moore's Law, for energy per bit function 10 8 Critical Dimension 10µm 1µm 100nm 10nm Gates including wires Gates only Technology Gap 10 7 10 6 10 5 10 4 10 3 10 2 10 Energy per Bit function (kt) 1 Transistor Measurements by Robert Chau, Intel 1960 1980 2000 Year 0.1 2020 2040 2060 Shoorideh and Yablonovitch, UCLA 2006