1

Embedded systems overview

What are they?

Design challenge optimizing design metrics Technologies

Processor technologies

IC technologies
Design technologies
2

Computing systems are everywhere Most of us think of desktop computers

PCs

Laptops
Mainframes Servers

But theres another type of computing system

Far more common...
3

Embedded computing systems

Computing systems embedded

Computers are in here... and here...

within electronic devices Hard to define. Nearly any computing system other than a desktop computer Billions of units produced yearly, versus millions of desktop units Perhaps 50 per household and per automobile

and even here...

Lots more of these, though they cost a lot less each.

Anti-lock brakes Auto-focus cameras Automatic teller machines Automatic toll systems Automatic transmission Avionic systems Battery chargers Camcorders Cell phones Cell-phone base stations Cordless phones Cruise control Curbside check-in systems Digital cameras Disk drives Electronic card readers Electronic instruments Electronic toys/games Factory control Fax machines Fingerprint identifiers Home security systems Life-support systems Medical testing systems

Modems MPEG decoders Network cards Network switches/routers On-board navigation Pagers Photocopiers Point-of-sale systems Portable video games Printers Satellite phones Scanners Smart ovens/dishwashers Speech recognizers Stereo systems Teleconferencing systems Televisions Temperature controllers Theft tracking systems TV set-top boxes VCRs, DVD players Video game consoles Video phones Washers and dryers

And the list goes on and on

5

Single-functioned
Executes a single program, repeatedly

Tightly-constrained
Low cost, low power, small, fast, etc.

Reactive and real-time

Continually reacts to changes in the systems

environment Must compute certain results in real-time without delay

Digital camera chip

CCD CCD preprocessor A2D lens JPEG codec Microcontroller Multiplier/Accum Pixel coprocessor D2A

DMA controller

Display ctrl

Memory controller

ISA bus interface

UART

LCD ctrl

Single-functioned -- always a digital camera Tightly-constrained -- Low cost, low power, small, fast Reactive and real-time -- only to a small extent

Obvious design goal:

Construct an implementation with desired

functionality

Key design challenge:

Simultaneously optimize numerous design

metrics

Design metric
A measurable feature of a systems

implementation Optimizing design metrics is a key challenge

8

Common metrics
Unit cost: the monetary cost of manufacturing each copy of
the system, excluding NRE cost

NRE cost (Non-Recurring Engineering cost):

The one-time monetary cost of designing the system

Size: the physical space required by the system Performance: the execution time or throughput of the
system

Power: the amount of power consumed by the system Flexibility:

the ability to change the functionality of the system without incurring heavy NRE cost
9

Common metrics (continued)

Time-to-prototype: the time needed to build a working
version of the system

Time-to-market: the time required to develop a system to

the point that it can be released and sold to customers initial release

Maintainability:

the ability to modify the system after its

Correctness, safety, many more

10

Power

Performance

Size

Expertise with both software and hardware is needed to optimize design metrics
Not just a hardware or

NRE cost

CCD lens

Digital camera chip A2D JPEG codec DMA controller CCD preprocessor Pixel coprocessor D2A

Microcontroller

Multiplier/Accum Display ctrl

software expert, as is common A designer must be comfortable with various technologies in order to choose the best for a given application and Hardware constraints
Software

Memory controller

ISA bus interface

UART

LCD ctrl

11

Revenues ($)

Time required to develop a product to the point it can be sold to customers Market window
Period during which the

product would have highest sales

Average time-toTime (months) market constraint is about 8 months Delays can be costly

12


Peak revenue Revenues ($)

Simplified revenue model

Product life = 2W, peak at

Peak revenue from delayed entry

On-time Market rise Delayed Market fall

W Time of market entry defines a triangle, representing market penetration Triangle area equals revenue

W Time

2W

Loss
The difference between the

On-time entry

Delayed entry

on-time and delayed triangle areas

13


Peak revenue Revenues ($)

Area = 1/2 * base * height

On-time = 1/2 * 2W * W
Delayed = 1/2 * (W-

Peak revenue from delayed entry

On-time Market rise Delayed Market fall

D+W)*(W-D)

Percentage revenue loss = (D(3W-D)/2W2)*100% Try some examples D=4 wks Lifetime 2W=52 wks, delay
(4*(3*26 4)/2*26^2) = 22% Lifetime 2W=52 wks, delay D=10 wks (10*(3*26 10)/2*26^2) = 50% Delays are costly!

W Time

2W

On-time entry

Delayed entry

14

Costs:

Unit cost: the monetary cost of manufacturing each copy of the system, excluding NRE cost NRE cost (Non-Recurring Engineering cost): The one-time monetary cost of designing the system total cost = NRE cost + unit cost * # of units per-product cost = total cost / # of units

= (NRE cost / # of units) + unit cost

Example
NRE=$2000, unit=$100 For 10 units
total cost = $2000 + 10*$100 = $3000 per-product cost = $2000/10 + $100 = $300
Amortizing NRE cost over the units results in an additional $200 per unit

15

Compare technologies by costs -- best depends on quantity

Technology A: NRE=$2,000, unit=$100 Technology B: NRE=$30,000, unit=$30 Technology C: NRE=$100,000, unit=$2
$200,000 $160,000 A B C $200 $160 A B C

tota l c ost (x1000)

$120,000

p er p rod uc t c ost

$120 $80

$80,000 $40,000 $0 0 800 1600 2400 Numb er of units (volume)

$40 $0 0 800 1600 2400 Numb er of units (volume)

But, must also consider time-to-market

16

Clock frequency, instructions per second not good measures Digital camera example a user cares about how fast it processes images, not clock speed or instructions per second

Widely-used measure of system, widely-abused

Latency (response time) Throughput

Time between task start and end e.g., Cameras A and B process images in 0.25 seconds
Tasks per second, e.g. Camera A processes 4 images per second Throughput can be more than latency seems to imply due to concurrency, e.g. Camera B may process 8 images per second (by capturing a new image while previous image is being stored).

Throughput speedup = 8/4 = 2

Speedup of B over S = Bs performance / As performance

17

Technology
A manner of accomplishing a task,

especially using technical processes, methods, or knowledge

Three key technologies for embedded systems

Processor technology
IC technology Design technology
18

The architecture of the computation engine used to implement a systems desired functionality Processor does not have to be programmable
Processor not equal to general-purpose processor
Controller Control logic and State register Datapath Register file General ALU Controller Control logic and State register Datapath Registers Custom ALU Controller Control logic State register Datapath

index
total +

IR

PC

IR

PC Data memory Data memory

Program memory Assembly code for:

Data memory

Program memory Assembly code for: total = 0 for i =1 to Application-specific Single-purpose (hardware)

total = 0 for i =1 to General-purpose (software)

19

Processors vary in their customization for the problem at hand

total = 0 for i = 1 to N loop total += M[i] end loop

Desired functionality

General-purpose processor

Application-specific processor

Single-purpose processor

20

Programmable device used in a variety of applications

Also known as microprocessor

Controller
Control logic and State register

Datapath
Register file

Features
Program memory General datapath with large register

IR

PC

General ALU

file and general ALU

User benefits
Low time-to-market and NRE costs High flexibility

Program memory
Assembly code for: total = 0 for i =1 to

Data memory

Pentium the most well-known, but there are hundreds of others

21

Digital circuit designed to execute exactly one program

a.k.a. coprocessor, accelerator or

Controller Control logic State register

Datapath index

peripheral

total
+

Features
Contains only the components needed

to execute a single program No program memory

Data memory

Benefits
Fast Low power Small size

22

Programmable processor optimized for a particular class of applications having common characteristics
Compromise between general-purpose

Controller Control logic and State register

Datapath Registers

and single-purpose processors

Custom ALU IR PC

Features
Program memory Optimized datapath Special functional units

Program memory
Assembly code for: total = 0 for i =1 to

Data memory

Benefits
Some flexibility, good performance, size

and power

23

The manner in which a digital (gate-level) implementation is mapped onto an IC

IC: Integrated circuit, or chip IC technologies differ in their customization to

a design Any processor can be mapped to any type of IC technology ICs consist of numerous layers (perhaps 10 or more)

24

Three types of IC technologies

Full-custom/VLSI
Semi-custom ASIC (gate array and standard

cell) PLD (Programmable Logic Device)

25

All layers are optimized for an embedded systems particular digital implementation
Placing transistors Sizing transistors Routing wires

Benefits
Excellent performance, small size, low power

Drawbacks
High NRE cost (e.g., $300k), long time-to-market

26

Lower layers are fully or partially built

Designers are left with routing of wires and

maybe placing some blocks

Benefits
Good performance, good size, less NRE cost

than a full-custom implementation (perhaps $10k to $100k)

Drawbacks
Still require weeks to months to develop
27

Designers can purchase an IC Connections on the IC are either created or

All layers already exist

destroyed to implement desired functionality PLA and PAL are simple PLD. FPGA is complex PLD.

Low NRE costs, almost instant IC availability Bigger, expensive (perhaps $30 per unit), power

Benefits

Drawbacks

hungry, slower

28

The most important trend in embedded systems

IC transistor capacity has doubled roughly every 18 months for the past several decades
10,000 1,000

Logic transistors per chip (in millions)

100 10 1 0.1

Note: logarithmic scale

0.01 0.001

29

1981 10,000 transistors

Leading edge chip in 1981

1984

1987

1990

1993

1996

1999

2002 150,000,000 transistors

Leading edge chip in 2002

This growth rate is hard to imagine, most people underestimate

A 2002 chip can hold about 15,000 1981 chips

Something that doubles frequently grows more quickly than most people realize!
inside itself

30

31

In January 1995, the Digital Alpha 21164

Wow!!

microprocessor had 9.3 million transistors. Six years later, a state of the art microprocessor contained more than 40 million transistors. It is theorised that with further miniaturisation, by 2015 these processors should contain more than 15 billion transistors, and by 2020 will be in molecular scale production, where each molecule can be individually positioned

32

The manner in which we convert our concept of desired system functionality into an implementation
Compilation/ Synthesis Libraries/ IP Hw/Sw/ OS Cores Test/ Verification Model simulat./ checkers Hw-Sw cosimulators

Compilation/Synthesis: Automates exploration and insertion of implementation details for lower level.
Libraries/IP: Incorporates predesigned implementation from lower abstraction level into higher level. Test/Verification: Ensures correct functionality at each level, thus reducing costly iterations between levels.

System specification

System synthesis Behavior synthesis

Behavioral specification

RT specification

RT synthesis

RT HDL simulators components

Logic specification

Logic synthesis

Gates/ Cells

Gate simulators
33

To final implementation


1983 1985 1987 1989 1991 1993 1995

1997

1999

2001

2003

2005

2007 2009

10

0.1

100

0.01

1,000

10,000

100,000

Exponential increase over the past few decades

34

Productivity (K) Trans./designer-month(K)

In the past:
Hardware and software

Sequential program code (e.g., C, VHDL) Compilers (1960's,1970's) Assembly instructions Assemblers, linkers (1950's, 1960's) Machine instructions Behavioral synthesis (1990's) Register transfers RT synthesis (1980's, 1990's)

design technologies were very different Recent maturation of synthesis enables a unified view of hardware and software

Logic equations / FSM's

Logic synthesis (1970's, 1980's) Logic gates

Hardware/software codesign

Microprocessor plus program bits: software

Implementation

VLSI, ASIC, or PLD implementation: hardware

The choice of hardware versus software for a particular function is simply a tradeoff among various design metrics, like performance, power, size, NRE cost, and especially flexibility; there is no fundamental difference between what hardware or software can implement.

35

Basic tradeoff
General vs. custom With respect to processor technology or IC technology The two technologies are independent

Implement multiple processors of different types on Singlethe same IC. GeneralASIP purpose purpose
General, providing improved:
Flexibility Maintainability NRE cost Time- to-prototype Time-to-market Cost (low volume) processor processor

Customized, providing improved:

Power efficiency Performance Size Cost (high volume)

PLD

Semi-custom

Full-custom

36

While designer productivity has grown at an impressive rate over the past decades, the rate of improvement has not kept pace with chip capacity
10,000 1,000 100,000 10,000 1000

Logic transistors per chip (in millions)

100 10 1 0.1 0.01 0.001

Gap

IC capacity

100 10 1

Productivity (K) Trans./Staff-Mo.

productivity

0.1 0.01

37

1981 leading edge chip required 100 designer months

10,000 transistors / 100 transistors/month

2002 leading edge chip requires 30,000 designer months

150,000,000 / 5000 transistors/month

Designer cost increase from $1M to $300M

10,000 1,000 100,000 10,000 1000 100 10 1 0.1 0.01

Logic transistors per chip (in millions)

100 10 1 0.1 0.01

Gap IC capacity productivity

Productivity (K) Trans./Staff-Mo.

0.001

38

The situation is even worse than the productivity gap indicates

In theory, adding designers to team reduces project completion time In reality, productivity per designer decreases due to complexities of team management and communication In the software community, known as the mythical man-month (Brooks 1975) At some point, can actually lengthen project completion time! (Too many cooks)
1M transistors, 1 designer=5000 trans/month Each additional designer reduces for 100 trans/month So 2 designers produce 4900 trans/month each
60000 50000 40000 30000 20000 10000 0 16 19 24 Months until completion 43 Individual 10 20 30 Number of designers 40 15 Team 16 18 23

39

Embedded systems are everywhere Key challenge: optimization of design metrics

Design metrics compete with one another

A unified view of hardware and software is necessary to improve productivity Three key technologies
Processor: general-purpose, application-specific, single-

purpose IC: Full-custom, semi-custom, PLD Design: Compilation/synthesis, libraries/IP, test/verification

40