Digital Signal Processor
Overview
Signal Processing
1D
ADC
0110111000101011010001
01010101110001010100111
10111010010001010100111
ADC
2D
Specialized
Real-time?
2
Real-time Processing
La definizione di Real-time Processing dipende dal
contesto applicativo
Real-time tasks
Example: 100-tap FIR filter è
eseguito real-time se si è in grado di
completare l’operazione di filtraggio entro
due istanti di capionamento successivi:
WT
PT
Waiting
Processing Time Time
99
y n    a k xn  k 
k 0
n
Sample Time
ST
n+1
WT  0
3
Analog Signal Processing
1D
2D
ASIC
4
Analog Signal Processing
Le tecniche di processing analocico si
avvalgono di componenti come:
Resistors, Capacitors e Inductors.
Le tolleranze associate a questi componenti
analogici, la temperatura, le variazioni di
voltaggio e le vibrazioni meccaniche
possono drammaticamente indurre effetti
non prevedibili sul circuito analogico.
5
Digital Signal Processing
1D
DSP
LW R1, a(R4)
Specialized
CPU
LW R2, b(R4)
ADD R3,R1,R2
SW c(R4),R3
ADC
0110111000101011010001
01010101110001010100111
10111010010001010100111
ADC
2D
FPGA
ASIC
Specialized HW
6
Digital Signal Processing
FIR Filter
Le attuali tecniche di processing digitale sono così
potenti che in particolari ambiti applicativi, è
quasi impossibile, per le tecniche di processing
analogiche ottenere performance comparabili.
FIR Filter with Linear Phase
Adaptive Filters
7
Digital Signal Processing
Programmabilità
Easy
Programmable
Hardware
DSP
FPGA
Change applications
Correct applications
Update applications
Cost
Reduces
Noise susceptibility
Development time
Power consumption
8
Digital Signal Processing
Limiti degli ADC
High frequency signals cannot be processed
digitally because of two reasons:
ADC
Analog to Digital Converters (ADC )
non lavorano ad alte frequenze
Le applicazioni diventano troppo complesse
per essere elaborate in Real-time
9
Digital Signal Processor
DSP vs. GPP
Digital Signal Processor (DSP)
Cost saving
Smaller size
Low power consumption
Processing of many “high” frequency
signals in real-time
General Purpose Processor (GPP)
Large memory
Advanced operating systems
10
Digital Signal Processor
Algoritmo SOP
La Somma di Prodotti (SOP) è la base
di molti algoritmi tipici per DSP:
Algorithm
Equation
M
Finite Impulse Response Filter
a
y ( n) 
k
x( n  k )
k 0
M
Infinite Impulse Response Filter

y (n) 
N
a k x ( n  k )
k 0
b
k
y (n  k )
k 1
N
Convolution
 x ( k ) h( n  k )
y ( n) 
k 0
N 1
Discrete Fourier Transform
X (k ) 
 x(n) exp[  j(2 / N )nk ]
n 0
Discrete Cosine Transform
F u  
N 1
 

 c(u). f ( x). cos 2 N u2 x  1
x 0
11
Digital Signal Processor
Hardware vs. Microcode
I processori DSP sono ottimizzati per operazioni
di somma e moltiplicazione. Moltiplicazioni
e addizioni sono eseguite con hardware
specializzati e completate in un ciclo di clock.
Example:
4-bit multiply
(unsigned)
Hardware
Microcode
1011
x 1110
1011
x 1110
10011010
0000
1011.
1011..
1011...
10011010
Cycle
Cycle
Cycle
Cycle
1
2
3
4
Cycle 5
12
Digital Signal Processor
Parametri Tecnici
Parameter
TMS320C6211
TMS320C6711
(@150MHz)
(@150MHz)
32-bit
32-bit
N/A
64-bit
Extended Arithmetic
40-bit
40-bit
Performance (peak)
1200MIPS
1200MFLOPS
2 (16 x 16-bit) with
32-bit result
2 (32 x 32-bit) with
32 or 64-bit result
32
32
Internal L1 program memory cache
32K
32K
Internal L1 data memory cache
32K
32K
Internal L2 cache
512K
512K
Arithmetic format
Extended floating point
Number of hardware multipliers
Number of registers
C6711 Datasheet: TMS320C6711.pdf
C6211 Datasheet: TMS320C6211.pdf
13
Digital Signal Processor
Parametri Tecnici
TMS320C6211
(@150MHz)
TMS320C6711
(@150MHz)
2 x 75Mbps
2 x 75Mbps
16
16
Not inherent
Not inherent
3.3V I/O, 1.8V Core
3.3V I/O, 1.8V Core
Yes
Yes
On-chip timers (number/width)
2 x 32-bit
2 x 32-bit
Cost
US$ 21.54
US$ 21.54
256 Pin BGA
256 Pin BGA
External memory interface controller
Yes
Yes
JTAG
Yes
Yes
Parameter
I/O bandwidth: Serial Ports
(number/speed)
DMA channels
Multiprocessor support
Supply voltage
Power management
Package
14
Digital Signal Processor
Fixed vs. Floating
High precision
Floating
Point
Ease of use
Wide dynamic range
High signal-to-noise ratio
It is the application that dictates which device
and platform to use in order to achieve
Can be
more expensive
optimum
performance
at a low cost
Can be slower than fixed-point
counterparts and larger in size
Higher power consumption
15
Digital Signal Processor
GP-DSP vs. ASIC-DSP
Application Specific Integrated Circuits
(ASICs) are semiconductors designed for
dedicated functions
Advantages
Disadvantages
•
•
•
•
•
•
•
•
•
High throughput
Lower silicon area
Lower power consumption
Improved reliability
Reduction in system noise
Low overall system cost
High investment cost
Less flexibility
Long time from design to
market
16
Digital Signal Processor
Texas Instruments’ TMS220 Family
C2000
Performance &
Efficiency
Best Ease-of-Use
Best MIPS per Watt /
- Multi Channel
Dollar /and
SizeMulti
Function App's
- Wireless
phones
-- Comm
Infrastructure
Internet
audio
players
Control
Systems
- Wireless Base-stations
- Digital
cameras
- Motorstill
Control
- DSL
-- Imaging
Modems
- Storage
-- Multi-media
Telephony
- Digital Ctrl
Systems
Servers
-- Video
VoIP
C5000
Lowest Cost
C6000
17
Digital Signal Processor
TMS220 - C6000
TMS320C62x: These first-generation fixed-point DSPs represent breakthrough
technology that enables new equipments and energizes existing implementations for
multi-channel, multi-function applications, such as wireless base stations, remote
access servers (RAS), digital subscriber loop (xDSL) systems, personalized home
security systems, advanced imaging/biometrics, industrial scanners, precision
instrumentation and multi-channel telephony systems.
TMS320C64x: The C64x fixed-point DSPs offer the industry's highest level of
performance to address the demands of the digital age. At clock rates of up to 1 GHz,
C64x DSPs can process information at rates up to 8000 MIPS with costs as low as
$19.95. In addition to a high clock rate, C64x DSPs can do more work each cycle with
built-in extensions. These extensions include new instructions to accelerate
performance in key application areas such as digital communications infrastructure
and video and image processing.
TMS320C67x: For designers of high-precision applications, C67x floating-point
DSPs offer the speed, precision, power savings and dynamic range to meet a wide
variety of design needs. These dynamic DSPs are the ideal solution for demanding
applications like audio, medical imaging, instrumentation and automotive.
18