Post on 21-Dec-2015
transcript
Welcome to Welcome to EE249: EE249: Embedded Embedded System DesignSystem DesignThe Real StoryThe Real Story
Alberto Sangiovanni-VincentelliAlberto Sangiovanni-VincentelliDepartment of EECS, University of California at Department of EECS, University of California at
BerkeleyBerkeley
2
AdministrationAdministration
Office hours: Office hours: Alberto’s : Tu-Th 12:30pm-2pm or (better) by Alberto’s : Tu-Th 12:30pm-2pm or (better) by
appointment (2-4882)appointment (2-4882)
Teaching Assistant: Teaching Assistant: Rong ChenRong Chen, rongchen@ic.eecs.berkeley.edu, rongchen@ic.eecs.berkeley.edu
3
GradingGrading
Grading will be assigned on:Grading will be assigned on:Homeworks (~30%) Homeworks (~30%)
Project (~50%)Project (~50%)
Reading assignments (~20%)Reading assignments (~20%)
There will be approx. 7 homeworks (due 2 weeks after There will be approx. 7 homeworks (due 2 weeks after
assignment) and 6 reading assignmentsassignment) and 6 reading assignments
4
Discussion sectionsDiscussion sections
Lab section (Th. 4-6):Lab section (Th. 4-6): tool presentationstool presentations
Discussion Session (Tu. 5-6)Discussion Session (Tu. 5-6) students’ presentation students’ presentation
of selected papersof selected papers Each student will have Each student will have
to turn in a one-paragraph report to turn in a one-paragraph report for each for each paper handed outpaper handed out
Each student (in groups of 2-3 Each student (in groups of 2-3 people) will have to make an oral people) will have to make an oral presentation once during the presentation once during the classclass
Auditors are OK but please Auditors are OK but please
register as P-NPregister as P-NP
Week Lab Sections Homeworks1 - - - - - -2 Tool presentation HW13 Discussion4 Tool presentation HW25 Discussion6 Tool presentation HW37 Discussion8 Tool presentation HW49 Discussion10 Tool presentation HW511 Discussion12 Tool presentation HW613 Discussion14 HW715
5
PlanPlan
We are on the edge of a revolution in the way electronics products are We are on the edge of a revolution in the way electronics products are
designeddesigned
System design is the keySystem design is the keyStart with the highest possible level of abstraction (e.g. control algorithms)Start with the highest possible level of abstraction (e.g. control algorithms)Establish properties at the right levelEstablish properties at the right levelUse formal modelsUse formal modelsLeverage multiple “scientific” disciplinesLeverage multiple “scientific” disciplines
Establish horizontal and vertical “supplier-chain” like partnershipsEstablish horizontal and vertical “supplier-chain” like partnerships
Need change in educationNeed change in education
6
Course overviewCourse overview
Managing Complexity
Orthogonalizing concerns
Behavior Vs.
Architecture
Computation Vs.
Communication
7
Behavior Vs. ArchitectureBehavior Vs. Architecture
SystemSystemBehaviorBehavior
SystemSystemArchitectureArchitecture
MappingMapping
Flow To ImplementationFlow To Implementation
CommunicationRefinement
BehaviorBehaviorSimulationSimulation
PerformancePerformanceSimulationSimulation
1
3
4
2
Models of Computatio
n
Performance models: Emb. SW, comm. and
comp. resources
HW/SW partitioning,Scheduling
SynthesisSW estimation
8
Behavior Vs. CommunicationBehavior Vs. Communication
Clear separation between functionality and interaction Clear separation between functionality and interaction
modelmodel
Maximize reuse in different environments, change only Maximize reuse in different environments, change only
interaction modelinteraction model
ETROPOLIS
PIG: Protocol interface generation
PEARLS: Latency insensitive protocols
9
Outline of the courseOutline of the course
Part 1. Introduction: Future of Information Technology, Part 1. Introduction: Future of Information Technology,
System Design, IP-based Design, System-on-Chip and System Design, IP-based Design, System-on-Chip and
Industrial TrendsIndustrial Trends
Part 2. Design Methodology (Platform-based Design, Part 2. Design Methodology (Platform-based Design,
Communication-based Design)Communication-based Design)
Part 3. Functional Design: Models of ComputationPart 3. Functional Design: Models of Computation
Part 4. Architecture Design: Capture, Exploration and MappingPart 4. Architecture Design: Capture, Exploration and Mapping
Part 5. Implementation Verification and Synthesis, Hardware Part 5. Implementation Verification and Synthesis, Hardware
and Softwareand Software
10
Introduction OutlineIntroduction Outline
Scenario and Characteristics of Future Information Scenario and Characteristics of Future Information TechnologyTechnology
Embedded Systems : Automotive, Home Networks, Smart Embedded Systems : Automotive, Home Networks, Smart Dusts, Universal RadiosDusts, Universal Radios
What is Needed at the Infrastructure LevelWhat is Needed at the Infrastructure Level
High-Leverage System Design Paradigms:High-Leverage System Design Paradigms:Communication-based DesignCommunication-based DesignArchitecture-Function Co-designArchitecture-Function Co-design
Platform-based Design as Implementation TechnologyPlatform-based Design as Implementation Technology
11
Electronics and the CarElectronics and the Car
•More than 30% of the cost of a car is now in Electronics•90% of all innovations will be based on electronic systems
12
Information Technology ScenarioInformation Technology Scenario
According to the International Data Corporation According to the International Data Corporation 96% of all Internet-access devices shipped in the United 96% of all Internet-access devices shipped in the United
States in 1997 were PCs. States in 1997 were PCs.
By the end of 2002, nearly 50% will not be PCs. Instead, they By the end of 2002, nearly 50% will not be PCs. Instead, they will be digital set-top boxes, cell phones, and personal will be digital set-top boxes, cell phones, and personal digital assistants, to name just a few. digital assistants, to name just a few.
By 2004, the unit shipments of such appliances will exceed By 2004, the unit shipments of such appliances will exceed those of the PC. those of the PC.
13
Historic PerspectiveHistoric Perspective
Technology discontinuities drive new computing Technology discontinuities drive new computing
paradigms and applicationsparadigms and applications
E.g., Xerox AltoE.g., Xerox Alto3Ms--1 mips, 1 megapixel, 1 mbps3Ms--1 mips, 1 megapixel, 1 mbps
Fourth M: 1 megabyte of memoryFourth M: 1 megabyte of memory
From time sharing to client-server with display intensive From time sharing to client-server with display intensive applicationsapplications
What will drive the next discontinuity? What are the new What will drive the next discontinuity? What are the new
metrics of system capability?metrics of system capability?
14
What’s Important: Shifts in Technology What’s Important: Shifts in Technology MetricsMetrics
Display (human-computer interface)Display (human-computer interface) More ubiquitous I/Os (e.g., MEMS sensors & actuators) and modalities More ubiquitous I/Os (e.g., MEMS sensors & actuators) and modalities
(speech, vision, image)(speech, vision, image)
How to Quantify?How to Quantify?
Connectivity (computer-computer interface)Connectivity (computer-computer interface) Not bandwidth but “scaled ubiquity”Not bandwidth but “scaled ubiquity”
Million accesses (wired and wireless) per dayMillion accesses (wired and wireless) per day
Computing (processing capacity)Computing (processing capacity) Unbounded capacity & utility functionality (very high mean time to Unbounded capacity & utility functionality (very high mean time to
unavailable, gracefully degraded capability acceptable)unavailable, gracefully degraded capability acceptable)
15
What’s Important: Shifts in User/Applications What’s Important: Shifts in User/Applications MetricsMetrics
Cost: Human EffortCost: Human EffortSave timeSave time
Reduce effortReduce effort
The Next Power ToolsThe Next Power ToolsLeveraging other peoples’ effort/expertiseLeveraging other peoples’ effort/expertise
e.g., “What did Dave read about disk prices?”e.g., “What did Dave read about disk prices?” e.g., “What did people who buy this book also buy?”e.g., “What did people who buy this book also buy?”
16
OutlineOutline
Scenario and Characteristics of Future Information Scenario and Characteristics of Future Information
TechnologyTechnology
Embedded Systems : Automotive, Home Networks, Smart Embedded Systems : Automotive, Home Networks, Smart
Dusts, Universal RadiosDusts, Universal Radios
What is Needed at the Infrastructure LevelWhat is Needed at the Infrastructure Level
High-Leverage System Design Paradigms:High-Leverage System Design Paradigms:Communication-based DesignCommunication-based Design
Architecture-Function Co-designArchitecture-Function Co-design
17
Chips Chips Everywhere!Everywhere!
CMOS Camera
SmartPen
Source: Dr. K. Pister, UC Berkeley
Chips that Fly?
18
Smart DustSmart Dust
Sensor
Interface
Power: battery, solar, cap.
Comm: LOS Optical (CCR, Laser)
Goal:
• Distributed sensor networks
• Sensor nodes:
•Autonomous
•1mm3
Challenges:
•1 Joule
•1 kilometer
•1 piece
19
Smart Dust ComponentsLaser diodeIII-V process
Passive CCR comm.MEMS/polysilicon
Active beam steering laser comm.MEMS/optical quality polysilicon
SensorMEMS/bulk, surface, ...
Analog I/O, DSP, ControlCOTS CMOS
Solar cellCMOS or III-V
Thick film batterySol/gel V2O5
Power capacitorMulti-layer ceramic
1-2 mm
20
Airborne Dust
Mapleseed solar cellMEMS/Hexsil/SOI
1-5 cm
Controlled auto-rotatorMEMS/Hexsil/SOIRocket dust
MEMS/Hexsil/SOI
22
Computing Revolution: Devices in the eXtremeComputing Revolution: Devices in the eXtreme
Evolution
Information Appliances:Scaled down desktops,e.g., CarPC, PdaPC, etc.
Evolved Desktops
Servers:Scaled-up Desktops,
Revolution
Information Appliances:Many computers per person,
MEMs, CCDs, LCDs, connectivity
Servers: Integrated withcomms infrastructure;Lots of computing in
small footprint
Display
Keyboard Disk
Mem
Proc
PC Evolution
Display Display
Camera
Smart
Sensors
Camera
Smart Spaces
ComputingRevolutionWAN
Server, Mem, Disk
InformationUtility
BANG!
Display
Mem
Disk
Proc
23
Modern Vehicles, an Electronic SystemModern Vehicles, an Electronic System
Electronic Toll CollectionElectronic Toll CollectionCollision AvoidanceCollision AvoidanceVehicle ID TrackingVehicle ID Tracking
Multiplexed SystemsMultiplexed Systems
VehicleVehicleCAN BusCAN Bus
BodyBodyControlControl
ECUECU ABSABS
SuspensionSuspension TransmissionTransmission
IVHS InfrastructureIVHS Infrastructure
Wireless Communications/DataWireless Communications/DataGlobal PositioningGlobal Positioning
Info/Comms/Info/Comms/AV BusAV Bus
CellularCellularPhonePhone
GPSGPS DisplayDisplay
NavigationNavigation Stereo/CDStereo/CD
SW Architecture
Network Design/Analysis Function / Protocol Validation
Performance Modelling
Supplier Chain Integration
24
Vehicles, a Consumer Electronic SystemVehicles, a Consumer Electronic System
CommsGSM/PCS
CDMA, PagingCompression
CommsGSM/PCS
CDMA, PagingCompression
S/W ShellWindows CE, NT, MAC, BIOS
S/W ShellWindows CE, NT, MAC, BIOS
S/W AppsBrowser,
Comms, User Apps
S/W AppsBrowser,
Comms, User Apps
ProcessorRISC, PowerPC
X86, Hitachi RISC
ProcessorRISC, PowerPC
X86, Hitachi RISC
DisplayHeads Up,Flat PanelGraphics
DisplayHeads Up,Flat PanelGraphics
User I/FVoice SynthesisVoice ControlStylus, ETC
User I/FVoice SynthesisVoice ControlStylus, ETC
Output & I/FSerial, Ethernet
Diagnostics
Output & I/FSerial, Ethernet
Diagnostics
Info/Comms/AV Bus
CellularPhone
GPS Display
Navigation Stereo/CD
• Minimum Technology to Satisfy User Requirement
• Usability• Integrate with Other
Vehicle Systems• Add the Function
Without Adding the Cost
Challenges
Vehicle Web SiteTechnology
25
When Will Dick Tracy’s Watch Be Available?When Will Dick Tracy’s Watch Be Available?
Ultimate Nomadic Tool in Broadband AgeUltimate Nomadic Tool in Broadband Age Two-way CommunicationTwo-way Communication
Language Translation & InterpretationLanguage Translation & Interpretation
e-Secretarye-Secretary
CameraCamera
MusicMusic
Electronic MoneyElectronic Money
26
Smart BuildingsSmart Buildings
•Task/ambient conditioning systems allow thermal conditioning in small, localized zones, to be individually controlled by building occupants , creating “micro-climates within a building”
• Other functions: security, identification and personalization, object tagging, seismic monitoring
Dense wireless network of Dense wireless network of sensor, monitor, and actuator nodessensor, monitor, and actuator nodes
•Disaster mitigation, traffic management and control• Integrated patient monitoring, diagnostics, and drug administration• Automated manufacturing and intelligent assembly• Toys, Interactive Musea
Dense wireless network of Dense wireless network of sensor, monitor, and actuator nodessensor, monitor, and actuator nodes
•Disaster mitigation, traffic management and control• Integrated patient monitoring, diagnostics, and drug administration• Automated manufacturing and intelligent assembly• Toys, Interactive Musea
27
PC/DataBased
PC-1
laptop
InternetAccess
PC-2
Printer
Telecom Based
VideoPhone
VoicePhone
PDA
Intercom
Appliance Based
Sprinklers
Toasters
Ovens
Clocks
ClimateControl
UtilityCustomization
Security Based
DoorSensorsMotion
Detectors WindowSensors
LightControl
AudioAlarms
Video surveillance
SmokeDetectors
Entertainment Based
Stereo
TV
Cam Corder
StillCamera
VideoGame
VCR
DVDPlayer
Web-TVSTB
Home Networking:Application (Subnet) Clusters
28
Silicon-Processed Micro-needlesSilicon-Processed Micro-needles
Lin and Pisano, IEEE/ASME J. of MEMS, Vol. 8, pp 78-84, 1999
• Neural probe with fluidchannel for bio-medical appl.
• Two micro-needles penetrating porterhouse (New-York) steak
29
Industrial Structure Shift
0
20
40
60
80
100
120
'98 '99 '00 '01 '02 '03
[M units]
PCPC
Game Machine
Game MachineCellular
Cellular DC0
50
100
1 2 3 4 5 6
100
50
0
'98 '00 '02
(%)
PC
DC
0
200
400
600
800
1000
1200
1400
1991 '93
'95
'97
'99
2001 '03
'05
'07
'09
'11
LSI Market Size( B$ )
SoC Market Size
World Wide Semiconductor Market Size
Market Structure Shift SOC Era has come.
•PC → DC•Wintel → Non-Wintel•Shift of Technology Driver
•Current Percentage of SoC Ratio is under 10%.⇒40 % in 2005, 70 ~ 80 % in 2010
•SoC is “single-seat constituency “, “take or not”.•Key Factor is the Synergy between Semiconductor & Set Divisions.
90‘s•PC
00‘s•High Performance ~ Game Machine•Low Power ~ Cellular
-Personal/Internet/Terminal
31
The Berkeley Wireless Research Center The Berkeley Wireless Research Center (BWRC)(BWRC)
Brodersen, Rabaey, Gray, Meyer, Katz, ASV, Tse and Brodersen, Rabaey, Gray, Meyer, Katz, ASV, Tse and
studentsstudents
Cadence, Ericsson, HP, Intel, Lucent, ST, TI, QualcommCadence, Ericsson, HP, Intel, Lucent, ST, TI, Qualcomm
Next Generation Wireless systems:Next Generation Wireless systems:CircuitsCircuits
ArchitecturesArchitectures
ProtocolsProtocols
Design MethodologiesDesign Methodologies
32
The “Universal” RadioThe “Universal” Radio
Fourth-generation radio providing following featuresFourth-generation radio providing following features
Focus on the wireless services with Focus on the wireless services with minimal constraintsminimal constraints on how the link is provided on how the link is provided
Allows for Allows for uncoordinated co-existenceuncoordinated co-existence of service providers (assuming they provide of service providers (assuming they provide
compatible services)compatible services)
Provides Provides evolving functionalityevolving functionality Adapts to provide requested service given type of service, location, and dynamic Adapts to provide requested service given type of service, location, and dynamic
variations in environment (i.e. number of users)variations in environment (i.e. number of users) Allows for to continuously upgrade to support new services as well as advances Allows for to continuously upgrade to support new services as well as advances
in communication engineering and implementation technologiesin communication engineering and implementation technologies
Presents an architectural vision to the multi-user, multi-service problem!Presents an architectural vision to the multi-user, multi-service problem!
This is in contrast with current approach where standards are the input and This is in contrast with current approach where standards are the input and
architecture the result - architecture the result - leading to spectral wastelandleading to spectral wasteland
33
Ultra Low-Power PicoRadioUltra Low-Power PicoRadio
Dedicated radio’s for ubiquitous wireless data acquisition and Dedicated radio’s for ubiquitous wireless data acquisition and
display. display.
Energy dissipation and footprint are of uttermost importanceEnergy dissipation and footprint are of uttermost importance
Goal: P < 1 mW enabling energy scavenging and self-poweringGoal: P < 1 mW enabling energy scavenging and self-powering
Challenges:Challenges: System architecture: self-configuring and fool-proofSystem architecture: self-configuring and fool-proof Ultra-low-power designUltra-low-power design Automated generation of application-specific radio modules making Automated generation of application-specific radio modules making
extensive use of parameterizable module generators and reusable extensive use of parameterizable module generators and reusable componentscomponents
34
Integrated CMOS RadioIntegrated CMOS Radio
AD
Analog RF
Timing recovery
phone
book
Java VM
ARQKeypad,Display
Control
FiltersAdaptive AntennaAlgorithms
Equalizers MUD
Accelerators(bit level)
analog digital
DSP core
uC core
(ARM)
Logic
Dedicated Logicand Memory
Integrate within the same chip very diverse system functions like:wireless channel control, signal processing, codec algorithms,
radio modems, RF transceivers… and implement them using a heterogeneous architecture
35
Communication versus ComputationCommunication versus Computation
Computation cost (2004): 60 pJ/operation (assuming continued Computation cost (2004): 60 pJ/operation (assuming continued
scaling)scaling)
Communication cost (minimum):Communication cost (minimum): 100 m distance: 20 nJ/bit @ 1.5 GHz100 m distance: 20 nJ/bit @ 1.5 GHz 10 m distance: 2 pJ/bit @ 1.5 GHz10 m distance: 2 pJ/bit @ 1.5 GHz
Computation versus CommunicationsComputation versus Communications 100 m distance: 300 operations == 1bit100 m distance: 300 operations == 1bit 10 m distance: 0.03 operation == 1bit10 m distance: 0.03 operation == 1bit
Computation/Communication requirements vary with distance, data Computation/Communication requirements vary with distance, data
type, and environmenttype, and environment
36
Energy-efficient Programmable Implementation PlatformEnergy-efficient Programmable Implementation Platform
EmbeddedMicroprocessor/
DSP System
ProgrammableLogic
DedicatedModules
ConfigurableArithmetic and Logic
Processors
Communication ChannelProtocol Processing
An
alog RF
“Software-defined Radio”
37
OutlineOutline
Scenario and Characteristics of Future Information Scenario and Characteristics of Future Information TechnologyTechnology
Embedded Systems : Automotive, Home Networks, Smart Embedded Systems : Automotive, Home Networks, Smart Dusts, Universal RadiosDusts, Universal Radios
What is Needed at the Infrastructure LevelWhat is Needed at the Infrastructure Level
High-Leverage System Design Paradigms:High-Leverage System Design Paradigms:Communication-based DesignCommunication-based DesignArchitecture-Function Co-designArchitecture-Function Co-design
Platform-based Design as Implementation TechnologyPlatform-based Design as Implementation Technology
38
What is Needed? What is Needed? (Endeavor Expedition,Berkeley, Oxygen, MIT)(Endeavor Expedition,Berkeley, Oxygen, MIT)
Automatic Self-ConfigurationAutomatic Self-Configuration Personalization on a Vast ScalePersonalization on a Vast Scale Plug-and-PlayPlug-and-Play
The OS of the PlanetThe OS of the Planet New management concerns: protection, information utility, not scheduling New management concerns: protection, information utility, not scheduling
the processorthe processor What is the OS of the Internet? TCP plus queue scheduling in routersWhat is the OS of the Internet? TCP plus queue scheduling in routers
Adapts to YouAdapts to You Protection, Organization, Preferences by ExampleProtection, Organization, Preferences by Example
39
Technology Changes & Architectural Technology Changes & Architectural Implications Implications
Zillions of Tiny DevicesZillions of Tiny Devices Proliferation of information Proliferation of information
appliances, MEMS, etc.appliances, MEMS, etc.
““Of course it’s connected!”Of course it’s connected!” Cheap, ample bandwidthCheap, ample bandwidth ““Always on” networkingAlways on” networking
Vast (Technical) CapacityVast (Technical) Capacity Scalable computing in the Scalable computing in the
infrastructureinfrastructure Rapid decline in processing, Rapid decline in processing,
memory, & storage costmemory, & storage cost
Adaptive Self-ConfigurationAdaptive Self-Configuration
Loosely OrganizedLoosely Organized
““Good Enough” Reliability and Good Enough” Reliability and
AvailabilityAvailability
Any-to-Any Transducers (dealing Any-to-Any Transducers (dealing
with heterogeneity, over time--with heterogeneity, over time--
legacy--and space)legacy--and space)
Communities (sharing)Communities (sharing)
40
Adaptive Self-ConfigurationAdaptive Self-Configuration
Plug-and-Play NetworkingPlug-and-Play Networking No single protocol/API: standardization processes too slow and stifle No single protocol/API: standardization processes too slow and stifle
innovationinnovation
Devices probDevices probee local environment and configure to inter-operate in that local environment and configure to inter-operate in that environmentenvironment
““Computer” not defined by the physical box: portals and ensemblesComputer” not defined by the physical box: portals and ensembles
Local Storage is a CacheLocal Storage is a Cache Invoke software and apps migrate to local diskInvoke software and apps migrate to local disk
System Learns Preferences by ObservationSystem Learns Preferences by Observation E.g., “Privacy by Example:” owner intervention on first access, observe E.g., “Privacy by Example:” owner intervention on first access, observe
and learn classification, reduce explicit intervention over timeand learn classification, reduce explicit intervention over time
41
Loose OrganizationLoose Organization
Loosely Structured InformationLoosely Structured InformationLarge volume, easily shared: supports communitiesLarge volume, easily shared: supports communities
Self-OrganizedSelf-OrganizedToo time consuming to do yourself: Organize by exampleToo time consuming to do yourself: Organize by example Individualized & context-Individualized & context-dependent filteringdependent filtering
Incremental Access, EventuallyIncremental Access, Eventually exact exactQuery by concept: “What did Dave read about storage prices?”Query by concept: “What did Dave read about storage prices?”
““A close answer quickly is better than a precise answer in the far future”; A close answer quickly is better than a precise answer in the far future”; Probabilistic access is often “good enough”Probabilistic access is often “good enough”
42
Any-to-Any TransducersAny-to-Any Transducers
No need for agreed upon/standardized APIs (though No need for agreed upon/standardized APIs (though
standard data types are useful)standard data types are useful) If applications cannot adapt, then generate transducers in the If applications cannot adapt, then generate transducers in the
infrastructure automaticallyinfrastructure automatically
Exploits compiler technologyExploits compiler technology
Enhance plug-and-play to the application levelEnhance plug-and-play to the application level
Legacy SupportLegacy Support Old file types and applications retained in the infrastructureOld file types and applications retained in the infrastructure
43
Next-Generation Operating EnvironmentsNext-Generation Operating Environments
Advances in hardware and networking will enable Advances in hardware and networking will enable an entirely an entirely
new kind of operating systemnew kind of operating system, which will raise the level of , which will raise the level of
abstraction significantly for users and developers.abstraction significantly for users and developers.
Such Such systemssystems will will enforce extreme location transparencyenforce extreme location transparency Any code fragment runs anywhereAny code fragment runs anywhere Any data object might live anywhereAny data object might live anywhere System manages locality, replication, and migration of computation and System manages locality, replication, and migration of computation and
datadata
Self-configuring, self-monitoring, self-tuning, scaleable and Self-configuring, self-monitoring, self-tuning, scaleable and
securesecure
Adapted from Microsoft “Millenium” White Paperhttp://www.research.microsoft.com
44
OutlineOutline
Scenario and Characteristics of Future Information Scenario and Characteristics of Future Information TechnologyTechnology
Embedded Systems : Automotive, Home Networks, Smart Embedded Systems : Automotive, Home Networks, Smart Dusts, Universal RadiosDusts, Universal Radios
What is Needed at the Infrastructure LevelWhat is Needed at the Infrastructure Level
High-Leverage System Design Paradigms:High-Leverage System Design Paradigms:Communication-based DesignCommunication-based DesignArchitecture-Function Co-designArchitecture-Function Co-design
Platform-based Design as Implementation TechnologyPlatform-based Design as Implementation Technology
46
System (for us)System (for us)
Environment to environmentEnvironment to environment
Sensors + Information Processing + ActuatorsSensors + Information Processing + ActuatorsComputer is a systemComputer is a system
Micro-processor is notMicro-processor is not
47
Embedded SystemsEmbedded Systems
Non User-ProgrammableNon User-Programmable
Based on programmable components (e.g. Micro-Based on programmable components (e.g. Micro-
controllers, DSPs….)controllers, DSPs….)
Reactive Real-Time Systems: Reactive Real-Time Systems: ““React” to external environmentReact” to external environment
Maintain permanent interactionMaintain permanent interaction
Ideally never terminateIdeally never terminate
Are subject to external timing constraints Are subject to external timing constraints (real-time)(real-time)
48
Electronic System Design Landscape:Electronic System Design Landscape:The Automotive CaseThe Automotive Case
Product DefinitionProduct Definition
IPIPDesign And AssemblyDesign And Assembly
ManufacturingManufacturing
Platforms
FabricsInterfaces
49
Disaggregation:Disaggregation:Complex Design Chain ManagementComplex Design Chain Management
Supply ChainSupply Chain
Movement of tangible goods from Movement of tangible goods from
sources to end marketsources to end market
Supply Chain Management is $3.8B Supply Chain Management is $3.8B
market projected to be $20B in 2005market projected to be $20B in 2005
Design ChainDesign Chain
Movement of technologyMovement of technology
(IP and knowledge) from sources to (IP and knowledge) from sources to
end marketend market
Design Chain Management is an Design Chain Management is an
untapped marketuntapped market
SemiconductorCompanies
SystemCompanies
Foundries System Test Equipment
Contract Manufacturing
Design Services
Embedded Software
Software Development Tools
EDA
Processor & Hardware IP
Process & Yield Services
Fabrication Equipment
IC Packaging & Test
Mechanical CAD
SubsystemCompanies
50
Supply Chain: Supply Chain: Design Roles-> Methodology->ToolsDesign Roles-> Methodology->Tools
Methodology
Design Roles
Tools
52
Product Specification & Architecture Definition(e.g., determination of Protocols and Communication standards)
System Partitioning and Subsystem Specification Critical Software Development System Integration
Automotive Supply Chain:Automotive Supply Chain:Car ManufacturersCar Manufacturers
53
1 Transmission ECU2 Actuation group3 Engine ECU4 DBW5 Active shift display6/7 Up/Down buttons 8 City mode button9 Up/Down lever10 Accelerator pedal
position sensor11 Brake switch
Subsystem Partitioning Subsystem IntegrationSoftware Design: Control Algorithms, Data ProcessingPhysical Implementation and Production
Automotive Supply Chain:Automotive Supply Chain:Subsystem ProvidersSubsystem Providers
54
HW layerHW layer
SW Platform layer(> 60% of total SW)SW Platform layer(> 60% of total SW)
Application Platform layer(10% of total SW)
Application Platform layer(10% of total SW)
Controllers Library
OSEKRTOS
OSEKCOMI/O drivers & handlers
(> 20 configurable modules)
Application Programming Interface
Boot LoaderSys. Config.
Transport
KWP 2000
CCP
ApplicationSpecificSoftware
Speedom
eterT
achometer
Water tem
p.
Speedom
eterT
achometer
Odom
eter---------------
ApplicationLibraries
Nec78kNec78k HC12HC12HC08HC08 H8S26H8S26 MB90MB90
CustomerLibraries
Automotive Supply Chain:Automotive Supply Chain:Subsystem ProvidersSubsystem Providers
Platform Integration: “firmware” and “glue software” Software Design: “Application”
55
HW layerHW layer
SW Platform layer(> 60% of total SW)SW Platform layer(> 60% of total SW)
Application Platform layer(10% of total SW)
Application Platform layer(10% of total SW)
Controllers Library
OSEKRTOS
OSEKCOMI/O drivers & handlers
(> 20 configurable modules)
Application Programming Interface
Boot LoaderSys. Config.
Transport
KWP 2000
CCP
ApplicationSpecificSoftware
Speedom
eterT
achometer
Water tem
p.
Speedom
eterT
achometer
Odom
eter---------------
ApplicationLibraries
Nec78kNec78k HC12HC12HC08HC08 H8S26H8S26 MB90MB90
CustomerLibraries
Automotive Supply Chain:Automotive Supply Chain:Platform & IP ProvidersPlatform & IP Providers
“Software” platform: RTOS and communication layer“Hardware” platform: Hardware and IO drivers
56
• Economics• Productivity• Process • IP Delivery & Reuse•Tools & Methodology•Manufacturing
How do we move SoC Design from the pilot line to production ?How do we move SoC Design from the pilot line to production ?
Source:M.Pinto, CTO, Agere
Issues Limiting SOC RampIssues Limiting SOC Ramp
57
SoC Landscape 2000+
• Total Cost Ownership• Average cost of a high end ASSP >$5M• Cost of fabrication and mask making has
increased significantly ($500k+ for masks alone)• SoC/ASIC companies look for a 5-10x return on development costs (~ $10M revenue)• Shorter and more uncertain product life cycles
• Compounding Complexities limiting Time-to-Market • Chip design complexity • Silicon process complexity • Context complexity • End-to-end verification
• New “System to Silicon” methodologies are required that recognized 80% of the system development is software
Source:M.Pinto, CTO, Agere
58
Productivity 2000+ ChallengeProductivity 2000+ Challenge
System ArchitectureSystem Architecture• HardwareHardware• SoftwareSoftware
Logic DesignLogic DesignVerificationVerification
Physical Physical DesignDesign
Silicon Silicon ProcessingProcessing
Will the design team deliver on time and within budget?Will the design team deliver on time and within budget?
15x - Productivity GAP15x - Productivity GAP
15M15M Tran
s./Staff-M
on
thT
rans./S
taff-Mo
nth
1.0M1.0M
100M 100M
10M10M
Lo
gic
Tra
ns.
/ Ch
ip
Lo
gic
Tra
ns.
/Ch
ip
( Av e
rag
e o
f T
op
1 0%
of
Co
de s
)(A
v era
ge
of
To
p1 0
% o
f C
od
e s)
Source:M.Pinto, CTO, AgereSource:M.Pinto, CTO, Agere
59
Process ChallengeProcess ChallengeCan you integrate what you need ?Can you integrate what you need ?
Lucent Modular Process StrategyLucent Modular Process Strategy• Communications focusCommunications focus• IP re-use across businessesIP re-use across businesses• Flexible system partitioningFlexible system partitioning• Only pay for what you needOnly pay for what you need• Leverage high volume platformLeverage high volume platform• Manufacture at fabs worldwideManufacture at fabs worldwide
MemoryMemory
ASICASICDSPDSP
Baseband
Baseband
Processing
Processing
High Frequency
High Frequency
(RF)(RF)
FiltersFilters
Mixers/VCOMixers/VCOLNA/PALNA/PA
High performance(speed, power, density)
core CMOS+SRAM platform
High performance(speed, power, density)
core CMOS+SRAM platform
Efficient (performance/cost)mix-and-match modules
Efficient (performance/cost)mix-and-match modules
++
Linear(to 4 masks)
RF(to 3 masks)
BiCMOS(3-4 masks)
FLASH(<4 masks)
FPGA/FPSC(1 mask)
eSRAM(1 mask)
“Fast Gate”(3 masks)
Copper(0 masks)
SiGeSiGe(4 masks)(4 masks)
Source:M.Pinto, CTO, AgereSource:M.Pinto, CTO, Agere
12/09/1999
60
Manufacturing Paradigm ChallengeInterconnection Dominates Fabrication Throughput
% of Fab of Interconnection vs. % of Fab Up-to-Contact
% o
f F
ab P
roce
ss
2LM 09µm
2LM 0.5 µm
3LM 0.35 µm
4LM 0.25 µm
Fab % up to contactFab % of interconnect
0
10
20
30
40
50
60
70
80
90
100
• Drives the need for new rapid prototype and production techniques• Impacts industry spare gate methodology for quick fixes • All metal programmable option lose their time to market advantage
• Drives the need for new rapid prototype and production techniques• Impacts industry spare gate methodology for quick fixes • All metal programmable option lose their time to market advantage
6LM 0.16 µm
Source:M.Pinto, CTO, Agere
61
Deep Submicron Paradigm ShiftDeep Submicron Paradigm Shift
40M Transistors2,000M Metal600 MHzWire RC 6 ns/cm
2M Transistors100M Metal100 MHzWire RC 1 ns/cm
200x1991 1996
Cell Based Design - Minimize Area - Maximize Performance - Optimize Gate Level
90%New
Design
Virtual Component Based Design - Minimize Design Time - Maximize IP Reuse - Optimize System Level 90%
ReusedDesign
22 22
62
Implementation Design TrendsImplementation Design Trends
Flat ASIC+
Platform BasedConsumerWirelessAutomotive
HierarchicalMicroprocessorsHigh end servers& W/S
Flat LayoutNet & Compute ServersBase stations
EDA
Flat ASIC
MicroP
63
Digital Wireless PlatformDigital Wireless Platform
AD
Analog RF
Timing recovery
phone
book
Java VM
ARQ
Keypad,Display
Control
FiltersAdaptive AntennaAlgorithm
s
Equalizers MUD
Accelerators(bit level)
analog digital
DSP core
uC core
(ARM)
Logic
Dedicated Logicand Memory
Source: Berkeley Wireless Research Center
64
Will the system solution match the original system spec?Will the system solution match the original system spec?
Software Hardware?
TxOptics
Synth/MUX
CDR/DeMUX
RxOptics
VCXO
mP
ClockSelect
LineI/F
OHPSTSPP
STSXC SPE
MapData
Framer
Cell/Packet
I/F
STMI/F
• IP Selection• Design• Verification
• Development• Verification• System Test
Concept
• Limited synergies between HW & SW teams• Long complex flows in which teams do not reconcile efforts until the end• High degree of risk that devices will be fully functional
• Limited synergies between HW & SW teams• Long complex flows in which teams do not reconcile efforts until the end• High degree of risk that devices will be fully functional
65
Historical EDA Focus
EDA Challenge to Close the Gap (SIA MARCO GSRC Project, EDA Challenge to Close the Gap (SIA MARCO GSRC Project, Berkeley Center) Berkeley Center)
Lev
el o
f A
bstr
acti
on
Behavior
SW/HW
RTL
Silicon
Concept to Reality
Gap
Design Entry Level
Gate Level “Platform”
Impact of Design Change(Effort/Cost)
Source: GSRC
• Industry averaging 2-3 iterations SoC design
• Need to identify design issues earlier
• Gap between concept and logical / Physical implementation
• Industry averaging 2-3 iterations SoC design
• Need to identify design issues earlier
• Gap between concept and logical / Physical implementation