Professional Documents
Culture Documents
Systems
1
Why Low Power
2
Why Low Power
• Increasing transistor count
• Higher speed of operation
• Greater device leakage currents
Increased process parameter variability due to
aggressive device size scaling has created
problems in yield, reliability, and testing.
3
Power Dissipation
4
Power Dissipation
• Power dissipation is measured commonly in terms of two
types of metrics:
1. Peak power: Peak power consumed by a particular device is
the highest amount of power it can consume at any time.
The high value of peak power is generally related to failures
like melting of some interconnections and power-line
glitches.
2. Average power: Average power consumed by a device is the
mean of the amount of power it consumes over a time
period. High values of average power lead to problems in
packaging and cooling of VLSI chips.
5
Sources of Power Dissipations
6
I1 is the reverse-biased p–n junction diode leakage current
I2 is the reverse-biased p–n junction current
I3 is the sub threshold leakage current
I4 is the oxide tunneling current(reduction due to oxide thickness)
I5 is the gate current(hot carrier injection)
I6 is the gate-induced drain leakage current 7
MOS Transistors
• The base semiconductor material used for the fabrication of
MOS integrated circuits is Silicon
Metal: Aluminum
Oxide: Sio2
Semiconductor : Silicon family
• Patterned layers of the three conducting materials: Metal, Poly
silicon, and diffusion.
• Fabrication involves a series of photolithographic techniques
and chemical processes involving oxidation of silicon, diffusion
of impurities into the silicon and deposition and etching of
aluminum, poly on the silicon to provide interconnection.
8
MOS Fabrication Technology
• Metal–oxide–semiconductor (MOS) fabrication is the
process used to create the integrated circuits (ICs)
• Wafer Fabrication
• Oxidation
• Mask Generation
• Photolithography
• Diffusion
• Deposition
9
CMOS Fabrication
• N-well Process
• P-Well Process
• Twin-Tub Process
• SOI Process
10
The n-Well Process
11
12
13
The p-Well Process
• P-well fabrication steps are similar to an n-well
process, except that a pwell is implanted to
form n-transistors rather than an n-well
14
Twin-Tub Process
In the twin-tub process, the starting material is either an
n+ or p+ substrate with a lightly doped epitaxial layer,
which is used for protection against latch-up. The
process is similar to the n-well process, involving the
following steps:
• Tub formation
• Thin oxide construction
• Source and drain implantations
• Contact cut definition
• Metallization
15
Twin-Tub Process
16
Short-Channel Effects
17
18
Channel Punch Through
19
20
21
22
Electrical Characteristics of MOS transistors
23
24
25
MOS Inverters
26
Transfer Characteristics and Noise Margins
27
MOS Inverter Configurations
• Pull up device as Resistor
• Pull up device as nMOS Depletion mode trans
• Pull up device as nMOS Enhancement mode
trans
• Pull up device as pMOS transistor
28
29
30
31
Comparison of the Inverters
32
Inverter Ratio in Different Situations
33
34
An nMOS Inverter Driven Through Pass
Transistors
35
36
Switching Characteristics
37
38
Driving large capacitance loads
• Super buffers
• BiCMOS inverters
• Buffer Sizing
39
Super buffers
40
Bi-CMOS inverters
41
MOS combinational circuits
• Pass Transistor Logic
• Gate Logic
• MOS Dynamic Circuits
42
Pass Transistor Logic
43
Realization of Pass transistor logic
44
CPL,SRPL & DPL
45
Gate Logic
46
MOS Dynamic Circuits
48
Two Phase dynamic circuits
No Static Power
No Ratioed Logic
49
50
Static /Dynamic
51
Charge leakage
52
Charge sharing
53
Clock Skew
54
Sources of Power Dissipation
• Short circuit power dissipation
• Switching power dissipation
• Glitching power dissipation
55
Introduction
• Short-circuit power: Short-circuit power dissipation occurs when both the
nMOS and pMOS networks are ON. This can arise due to slow rise and
fall times of the inputs .
• Switching power dissipation: As the input and output values keep on
changing, capacitive loads at different circuit points are charged and
discharged, leading to power dissipation. This is known as switching
power dissipation. Until recently, this was the most dominant source of
power dissipation.
• Glitching power dissipation: Due to a finite delay of the logic gates, there are
spurious transitions at different nodes in the circuit. Apart from the
abnormal behavior of the circuits, these transitions also result in power
dissipation known as glitching power dissipation.
56
Introduction
57
Short-Circuit Power Dissipation
58
59
Switching Power Dissipation
60
61
Switching Activity of Static CMOS Gates
62
Inputs Not Equiprobable
63
Switching Activity for complex logic gates
64
Glitching Power Dissipation
Glitching power dissipation: Due to a finite delay of the logic gates, there are spurious
transitions at different nodes in the circuit. Apart from the abnormal behavior of the
circuits, these transitions also result in power dissipation known as glitching power
dissipation.
65
Leakage Power Dissipation
When the circuit is not in an active mode of operation, there is static power
dissipation due to various leakage mechanisms
Leakage’s are :
66
67
p–n Junction Reverse-Biased Current
68
Band-to-Band Tunneling Current
69
Sub threshold Leakage Current
• The sub threshold leakage current in CMOS circuits is due to carrier diffusion
between the source and the drain regions of the transistor in weak inversion
70
Vth roll off
71
Tunneling Through Gate Oxide
72
Hot – carrier injection
73
Supply Voltage Scaling for Low Power
74
75
• Static Voltage Scaling (SVS) In this case, fixed supply voltages
are applied to one or more subsystems or blocks.
• Multilevel Voltage Scaling (MVS) This is an extension of the
SVS, where two or few fixed discrete voltages are applied to
different blocks or subsystems.
• Dynamic Voltage and Frequency Scaling (DVFS) This is an
extension of the MVS, where a large number of discrete
voltages are applied in response to the changing workload
conditions of the subsystems.
• Adaptive Voltage Scaling (AVS) This is an extension of the
DVFS, where a close loop control system continuously
monitors the workload and adjusts the supply voltage.
76
77
78
Constant-Field Scaling
79
Constant-Voltage Scaling
80
Short-Channel Effects
81
Architectural-Level Approaches
• Architectural level to Register-transfer-Level.
3 level of approach
82
Parallelism for low power
• Parallel processing is traditionally used for the
improvement of performance at the expense
of a larger chip area and higher power
dissipation
83
84
85
Pipelining
86
87
Combining Parallelism with Pipelining
88
89
Voltage Scaling Using High-Level
Transformations
90
Loop unrolling
91
92
93
94
95
Voltage scaling interfaces/ converter
placement
96
97
• Floor Planning, Routing, and Placement
• Static timing analysis
• Power up and power down sequencing
• Clock frequency
98
99
Dynamic Voltage and Frequency Scaling
100
101
102
103
Variable Voltage Generator V(r)
104
105
Adaptive Voltage Scaling
106
107
108
109