Embedded Design

and Architecture
Tien Pham Van, Dr. rer. nat.
Hanoi University of Science and Technology
Compiled with reference to other presentations

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Outlines

• Specification
• System architecture
• Co-design
• Hardware components
• Modeling languages and tools

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Parts of an Embedded System

USER EMBEDDED SYSTEM

I/O

MEMORY PROCESSOR

HARDWIRED UNIT
• Application-specific logic
• Timers
• A/D and D/A conversion

ENVIRONMENT
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Design flow: inside a tool

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Co-design: Complex Systems

• Abstraction
– Reduce the number of objects managed by a
design task, e.g., by grouping objects using
hierarchy
– Computer-aided design (CAD) example
• Logic level: transistors grouped into gates
• Register transfer level (RTL): gates grouped into
registers, ALUs, and other RTL components

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Generic Co-Design Methodology
Synthesis model
• Specification Analysis
• Allocation task &
• Partitioning Validation
• Scheduling
Note: design
models may
• Communication be captured
synthesis in the same
language
Implementation
• Software synthesis
• Hardware synthesis
• Interface synthesis

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 • Hardware and Software Co Design - System
Integration - Embedded System & RTOS -
YouTube
https://www.youtube.com/watch?v=w-KHW7ZxMW8

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 System Specification

• Describes the functionality of the system

without specifying the implementation
• Describes non-functional properties such as
performance, power, cost, and other quality
metrics or design constraints
• May be executable to allow dynamic
verification

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 System Specification Example

B0: top
shared sync
behavior
• integer read child
write
variable behavior
• boolean
variable
sync
Graphical
representation:
• Hierarchy
• Concurrency Behaviors
• Transitions • Sequential: B1, B2, B3
between • Concurrent: B4, B5
behaviors • Atomic: B1
Embedded Networking Research Group
• Composite: B2
School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 System Specification Example (cont.)

Producer- shared sync

consumer
functionality write read

• B6 computes a
value
• B4 consumes the
value sync

• Synchronization
is needed: B4
waits until B6
produces the
value

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 System Specification Example
• Atomic behaviors
B1( ) B3( ) B7( )
{ { {
stmt; stmt; stmt;
... ... ...
} } }

B6( ) B4( )
{ {
int local; int local;
… wait(sync);
shared = local + 1; local = shared - 1;
signal(sync); ...
} }

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Allocation

• Selects the type and number of components

from a library and determines their
interconnection
• Implements functionality so as to
– Satisfy constraints
– Minimize objective cost function
• Result may be customization of a generic
target architecture

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Allocation Example

PE1 PE2

Proc1 LMem1 GMem1 ASIC1 LMem2

bus1 bus2 bus3

IF1 IF2 IF3

system bus

PE: Processing Element Arbiter1

LMem: Local Memory Target Architecture Model

GMem: Global Memory
IF: Interface

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Partitioning

• Defines the mapping between the set of

behaviors in the specification and the set of
allocated components in the architecture
– Satisfy constraints
– Minimize costs
• Not yet near implementation
– Multiple behaviors in a single PE (scheduling)
– Interactions between PEs (communication)
• Design model
– Additional level of hierarchy
– Functional equivalence with specification
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Partitioning Example
synchronization variables
Top shared sync B1_start B1_done B4_start B4_done

PE0 PE1
controlling Child (B1)
B1_start
behavior B1_ctrl assigned to
B1_done different PE
than
B4_start parent (B0)
B4_ctrl

B4_done

System model after partitioning

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 • Hardware/Software Partitioning - 1 - YouTube
https://www.youtube.com/watch?v=Gkp753foAgE

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Scheduling

• Given a set of behaviors and optionally a set

of performance constraints, determines a
total order in time for invoking behaviors
running on the same PE
• Maintains the partial order imposed by
dependencies in the functionality
• Minimizes synchronization overhead
between PEs and context-switching overhead
within each PE

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Scheduling

• Ordering information
– Known at compile time
• Static scheduling
• Higher inter-PE synchronization overhead if inaccurate
performance estimation, i.e., longer wait times and
lower CPU utilization
– Unknown until runtime (e.g., data-, event-
dependent)
• Dynamic scheduling
• Higher context-switching overhead (running task
blocked, new task scheduled)

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Scheduling Example

shared sync B6_start B3_start

Scheduling
decision:
• Sequential
B6_start
ordering of
behaviors
sync
on PE0, PE1
• Synchronization
B3_start
to maintain
partial order
across PEs
• Optimization - no
control behaviors
System model after static scheduling

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Scheduling Example (cont.)
• Atomic behaviors
B1( ) B3( ) B7( )
{ { {
… wait(B3_start); stmt;
signal(B6_start); ... ...
} } }

B6( ) B4( )
{ {
int local; int local;
wait(B6_start); wait(sync);
… local = shared - 1;
shared = local + 1; …
signal(sync); signal(B3_start);
} }

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communication Synthesis

• Implements the shared-variable accesses

between concurrent behaviors using an inter-
PE communication scheme
– Shared memory: read or write to a shared-
memory address
– Local PE memory: send or receive message-
passing calls
• Inserts interfaces to communication channels
(local or system buses)

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communication example

lbus0 lbus1 lbus2 sbus

Synthesis IF0 IF1 IF2

decision:
PE0 PE1
• Put all
global B6

Shared_mem
B1
variables

Arbiter
into B7
Shared_mem
B4
• New global B3

variables in Top

System model after communication synthesis

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communication Example (cont.)

• Atomic behaviors
B1( ) B3( ) B7( )
{ { {
… wait(*B3_start_addr); stmt;
signal ... ...
(*B6_start_addr); }
}
}

B6( ) B4( )
{ {
int local; int local;
wait (*B6_start_addr); wait (*sync_addr);
… local = *shared_addr - 1;
*shared_addr = local + 1; …
signal(*sync_addr); signal (*B3_start_addr);
} }

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communication Example (cont.)
• Atomic behaviors
IF0( ) IF1( ) IF2( ) Arbiter( )
{ { { {
stmt; stmt; stmt; stmt;
... ... ... ...
} } } }

Shared_mem( )
{
int shared;
bool sync;
bool B3_start;
bool B6_start;
}

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Implementations

– Processor: compiler translates model into

machine code
– ASIC: high-level synthesis tool translates model
into netlist of RTL components
– Interface
• Special type of ASIC that links a PE with other
components
• Implements the behavior of a communication channel

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 RTL

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 High level synthesis

• https://youtu.be/gTuWkoOq1k0?t=481

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Case study – Xilinx HLS

• Introduction to Vitis High-Level Synthesis

(HLS) - YouTube

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Hardware Component Design
Reference to “Embedded System Design”, by
Peter Marwedel

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Overview

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Block diagram

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Inputs

• Acceleration sensors (inertial measurement

units): accelerometers, gyros
• Image sensors: CCCD, CMOS
• Biometric sensors
• Artificial eyes
• RFID
• Automotive sensors

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Processing unit – hardware tech
• ASIC: high energy-efficient, high cost of designing
and manufacturing, less flexible
– MPW (multi-project wafer) may decrease a bit, but hard
to correct errors
• Reconfigurable logic: FPGA, CLPD
• Processors and controllers
– ARM, MIPS, x86/ATOM
– Multi-core
– DSP
– VLIW
– GPU
– MPSoC
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Processors

• Flexibly programmable
• Less energy-efficiency than ASIC
• Off-the-shelf (COTS)
• Diverse (as presented earlier) to meet various
applications
• Well-know families: X86/ATOM, ARM, MIPS,
PowerPC, DSP

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 • Types of Processor, Units of Processor,
Classifications of Processor, Embedded
System - YouTube
https://www.youtube.com/watch?v=obknO3gA92E

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 ARM case study

• Comparison with CISC:

https://youtu.be/AADZo73yrq4?t=639
• Architecture details: ARM Processor
Fundamentals – YouTube
https://www.youtube.com/watch?v=8028NpYig8s
• Wiki: ARM architecture - Wikipedia
https://en.wikipedia.org/wiki/ARM_architecture_famil
y

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Energy efficiency

• CMOS circuits are preferred

• Dynamic power consumption (associated
with switching) + Static power consumption
• Measures to improve efficiency
– Parallel execution
– Dynamic power management (DPM)
– Dynamic voltage and frequency scaling (DVFS)

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Code size efficiency - compression

• ARM design:

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Execution time efficiency
• To reduce execution time without increasing clock speed.
Output element xs corresponds to a weighted average over the last n signal
elements of w and can be computed iteratively, adding one product at a time.
Processors for DSP are designed such that each iteration can be encoded as a
single instruction.

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Very long instruction word – VLIW processors

• Instructions explicitly identify operations to be performed in

parallel. This is possible with explicit parallelism instruction
set computers (EPICs)
• Detection of parallelism is moved from the processor to the
compiler, which saves energy at run-time
• Several operations or instructions (called fields) are encoded
in a long instruction word and are assumed to be executed in
parallel. Each field controls certain hardware units

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Multi-core
• Multiple cores are integrated into a single die
• Cache: L1 is private, L2 is shared
• Cores can be homo- or heterogeneous

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Graphics Processing Unit - GPU
• To run a large number of computations concurrently in order to achieve
the desired performance.
• GPU runs many fine-grained threads at the same time. If some thread
gets blocked, e.g., due to waiting for memory, the core will execute some
other thread.
• The instructions contained in a thread can also be executed concurrently,
e.g., by using multiple pipelines.
SIMD: single instruction
multiple data

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 MPSoC (multiprocessor SoC)

• Heterogeneous core systems also contain GPU, DSP

• Additional modules such as MMU, interfaces for peripherals,
etc may be integrated into a single die
• Mapping is critical for energy efficiency

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Microcontrollers

• Small computer that contains processors, memory, and I/O

peripherals
• Mixed signal microcontrollers are common, integrating
analog components needed to control non-digital electronic
systems
• They generally retain functionality while waiting for an event
such as a button press or other interrupts
• Power consumption may be very low thanks to flexible
control of component states

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Microcontrollers - components
• central processing unit: 4-bit processors to complex 32-bit or 64-bit
• volatile memory (RAM) for data storage
• ROM, EPROM, EEPROM or Flash memory for program and operating
parameter storage
• discrete input and output bits, allowing control or detection of the logic
state of an individual package pin
• serial input/output such as serial ports (UARTs)
• other serial communications interfaces like I²C, Serial Peripheral
Interface and Controller Area Network for system interconnect
• peripherals such as timers, event counters, PWM generators, and
watchdog
• clock generator – often an oscillator for a quartz timing crystal, resonator
or RC circuit
• many include analog-to-digital converters, some include digital-to-analog
converters
• in-circuit programming and in-circuit debugging support
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Reconfigurable logic
• Harmonized between slow software solutions and expensive
ASIC
• FPGA is the most popular, beside CPLD
• Be suitable for:
– Fast prototyping
– Low-volume apps
– Real-time systems
– Enhanced level of parallel processing

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 A close-up view

CLB
Switchbox Routing
Channel
IOB

Channel
Routing
Configuration
Bit

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Configurable logic

• Configurations (generated from VHDL, Verilog) define both

hardware and software operations, and can be stored
persistently in ROM, flash, or temporally in RAM
• Integrated processors:
– Hardcore: ZynQ ultrascale MPSoC, which contains ARM cortex cores
– Softcore: Microblaze

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Memory

• Working memory: DRAM

• Registers: very fast, but limited
– RISC processors: possibly hold both opcode and operands
• Persistent memory: flash, hard disk, cloud

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Memory - cache

• Cache, TLB (translation look-aside buffer), SPM (scratchpad

memory): to resolves speed difference between RAM and
registers
– Cache: use of tags (subset of address bits) to look for a valid copy;
parallelly reading multiple entries before selecting the right one; levels
L1, L2 and possibly L3 integrated into processor
– TLB: for virtual memory management; associated with MMU
– SPM: to access when simple address decoder signals an address to be in
the range of SPM; can be in the same die with processor (hence called
on-chip memory)

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Cache

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communication

• Means/media: wireless (radio, infrared), optical (fibers), and

wires
• Requirements:
– Real-time constraint: TDMA vs CSMA/CA
– Efficiency: hardware utilization, wires, energy
– Bandwidth
– Latency
– Event-driven
– Security/privacy
– Safety/robustness
– Fault-tolerance
– Maintainability/diagnosability

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Typical communication means
• Sensor/actuator bus
• Field bus:
– Controller Area Network (CAN): inexpensive connectivity between
controllers and peripherals; widely used in automotive industry and
later in smart home; based on differential signaling and arbitration
using CSMA/CA (hence starvation possibly occurs)
– Time-Triggered Protocol (TTP): This is a protocol for fault-tolerant
safety systems like airbags in cars
– FlexRay™: a TDMA protocol developed by the FlexRay consortium
(BMW, Daimler AG, GM, Ford, Bosch, Motorola, and Philips)
– LIN (Local Interconnect Network): This is a low-cost communication
standard for connecting sensors and actuators in the automotive
domain
– MAP: MAP is a bus designed for car factories
– EIB: The European Installation Bus (EIB) is a bus designed for smart
homes.
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Typical communication means
• The Inter-Integrated Circuit (I2C) Bus : This is a simple low-
cost bus designed to communicate at short distances (meter
range) with relatively low data rates. The bus needs only four
wires: ground, SCL (clock), SDA (data), and a voltage supply
line
• Wired multimedia communication: For wired multimedia
communication, larger data rates are required. For example,
MOST (Media Oriented Systems Transport) is a
communication standard for multimedia and infotainment
equipment in the automotive domain. Standards like IEEE
1394 (FireWire) may be used for the same purpose
• Wireless communication: This kind of communication is
becoming more popular. Popular interfaces include cellular,
LoRA, Bluetooth, ZigBee, etc
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Outputs
• Display: Organic displays are emitting light
and can be fabricated with very high
densities. In contrast to LCDs, they do not
need backlight and polarizing filters. Major
changes are therefore expected in these
markets
• Electro-mechanical devices: These influence
the environment through motors and other
electro-mechanical equipment
– Actuators can be a huge motors to microsystems
• Key techniques:
– DAC: not easy to integrate into digital chips
– PWM: output voltage depends on values stored in
registers

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Energy matters

• Source: mainly battery powered; energy harvesting has

been emerging (photovoltaic, piezoelectric,
thermoelectric generators, kinetic energy, ambient
electromagnetic radiation )
• Storage: non-rechargeable/rechargeable battery,
capacitor
• Energy-efficiency:
– Appropriately controlling hardware operations
– Properly selecting hardware tech: FPGA, ASIC, DSP,
Processors…
– Programming differently (unlike many HUST students…)

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Embedded Modeling: tools
and languages
Reference to “Embedded System Design”,
by Peter Marwedel

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Model concept

• In embedded world, system under design (SUD) must

be modeled for researching specifications
• “A model is a simplification of another entity, which can
be a physical thing or another model. The model
contains exactly those characteristics and properties of
the modeled entity that are relevant for a given task. A
model is minimal with respect to a task if it does not
contain any other characteristics than those relevant
for the task”
• Models are described in modeling languages, which
describes system and component specifications

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Required capabilities of modeling languages
• Behavioral and structural hierarchy
• Component-based design
• Concurrency
• Synchronization and communication
• Timing behaviors
• State/exception-oriented behavior von Neumann MoC is
composed of tasks,
• Presence of programming elements process and threads
• Executability with communications,
which has no timing
• Portability and flexibility description
• Termination
• Support for the design of dependable systems (e.g. formal
verification)
• Appropriate model of computation (MoC) – not von Neumann MoC
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 MoC
• MoC defines:
– components and their organization of computations, e.g. procedures,
processes, functions, and finite state machines
– communication protocols: describe methods for communication between
components, e.g. asynchronous message passing and rendezvous-based
communication
• Relations between components are represented in task graphs
and process networks
– Nodes represent computations
– Edges represent relations between components: causal dependence,
timing, information exchange, etc

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 MoC - what task graph includes
• Timing information: arrival time, execution time, deadline…
• Types of relations between computations: precedence, I/O, etc
• Exclusive access to resources
• Periodic schedules, e.g. DSP
• Hierarchical graph nodes

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 MoC – communication fashions
• Shared memory
– Fast but difficult in multiprocessor system without common physical
memory
– Critical sections must be identified
• Message passing
– Loosely coupled processors
– Three kinds: blocking, non-blocking, and extended rendezvous/remote
invocation

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 MoCs and languages

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Early phase design – Use Case

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Early phase design – Sequence Charts and
Time/Distance Diagrams

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Early phase design – Differential Equation

• Modeling a bouncing ball

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communicating Finite State Machines (CFSMs) -
StateCharts – for shared mem. comm.

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Communicating Finite State Machines (CFSMs) -
SDL for message-based com.

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Data Flow – Kahn Process Network

• A distributed model of computation in which a group

of sequential processes communicate through FIFO channels.
• Reading from a channel is blocking while writing is non-
blocking

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Data Flow – SDF (static data flow)
• A directed graph that facilitates scheduling and buffer calculation
– Node: actor
– Edge: token

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Data Flow –Simulink
• Computational graph structures are also frequently used in
control engineering.
– An approach of model-based design
– Analog blocks included
– Ideal timing (fast) model is assumed
– Code generation for various language implementation: C, VHDL...

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Control Flow – Petri Nets
• Very comprehensive descriptions of control flow. Conditions,
events, and a flow relation are the key elements. An event is
associated with pre- and post-condition

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Activity Diagram

• A UML diagram that

includes extended
Petri nets

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Discrete event-based languages - TLM

• Only care about events, not modeling continuous

[physical] time
• Separation between computation and
communication
• In TLM (transaction-level modeling), the emphasis
is more on the functionality of the data transfers
— what data are transferred to and from what
locations — and less on their actual
implementation, that is, on the actual protocol
used for data transfer

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Discrete event-based languages - TLM
• A: untimed model/specification model; B: component assembly model
• C: bus arbitration model; D: bus functional model
• E: cycle-accurate computation model; F: implementation model

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 TLM – Spec C

• Spec C models systems as

hierarchical networks of
behaviors communicating
through channels.
• Spec C descriptions consist of
behaviors, channels, and
interfaces

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 TLM – SystemC
• TLM modeling and the separation
between communication and
computation are also available in
SystemC™. Similar to SpecC, SystemC
provides channels, ports, and
interfaces as abstract components for
communication
• System C comprises a notion of
processes executed concurrently.
P&R: Place & Route
Their execution is controlled by calls to
wait primitives and sensitivity lists
(lists of signals for which value
changes start a re-execution of code)
• Hardware synthesis starting from
SystemC has become available
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 SystemC

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Discrete Event-Based Languages – VHDL
• VHDL is another HDL which is based on the discrete event paradigm.
Unfortunately, it does not support a clear distinction between
communication and computation, and reusing components is more
difficult

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Discrete Event-Based Languages – [System]Verilog
• Much like VHDL. Verilog versions 3.0 and 3.1 are also known as
SystemVerilog. They include numerous extensions to Verilog 2.0:
– additional language elements for modeling behavior,
– C data types such as int and type definition facilities such as typedef and
struct,
– definition of interfaces of hardware components as separate entities,
– standardized mechanism for calling C/C++ functions and, to some extent, to
call built-in Verilog functions from C,
– significantly enhanced features for describing an environment (called test
bench) for the hardware circuit under design (called CUD), and for using the
test bench to validate the CUD by simulation,
– classes known from object-oriented programming for use within test
benches,
– dynamic process creation, standardized interprocess communication and
synchronization, including semaphores,
– automatic memory allocation and deallocation,
– language features that provide a standardized interface to formal
verification
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 von Neumann languages
• CSP (communicating sequential processes) is one of the first languages
comprising mechanisms for interprocess communication
• ADA is the language to address dependability and maintainability, with
all software written in the same real-time programming

• Communication libraries:
– MPI (message passing interface)
– OpenMP

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Levels of Hardware Modeling
• System-level models: represents both embedded system and the
environment. Examples are VHDL-AMS, Verilog-AMS SystemC-AMS, etc
• Algorithmic level: simulates algorithms run on the system, e.g. H26x
• Instruction set level: works on compiled algorithm, counts instructions
execution time
• RTL level: models all the RTL components, including arithmetic/logic units
(ALUs), registers, memories, multiplexers, and decoders
• Gate level: provides more accurate information about signal transition
probabilities, power estimations and delay
• Circuit model: works on circuit theory and its components e.g. current and
voltage sources, resistors, capacitances, inductances, etc. Simulations
involve partial differential equations
• Layout models: reflects the actual circuit layout. Such models include
geometric information
• Process and device models: represents model fabrication process

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Comparison of languages and tools

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Combination of multiple languages and tools

Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Unified Modeling Language - UML
• A language includes diagrams representing many models of
computations

• Various diagrams of deployment, package, class,

communication, component, object, interaction overview, and
composite structure are provided
• Combination of UML and others such as SDL, C++ is possible
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596
 Review

• Workflows of specification, co-design and

synthesis
• Detailed implementation of partitioning,
allocation, scheduling, and communication
• Systematically understanding hardware
architecture
• How to work on modeling languages and
tools
 Able to take over assignment projects!
Embedded Networking Research Group School of Elec. and Telecom - Hanoi University of Science and Technology
Email: [email protected] C9-411, Dai Co Viet str. 1, HBT, Hanoi Tel: +84-243-8693596