Power. Motivation for design constraints of power consumption Power metrics Power consumption analysis in CMOS How can a logic designer control power?. 500 Million. Internet Computers. Today’s Internet. Internet Users. 1.5 Billion. Automobiles 700 Million. Telephones 4 Billion.
Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.
CS 150  Spring 2007 – Lec #28 – Power  1
Million
Internet Computers
Today’s Internet
Internet Users
1.5 Billion
Automobiles
700 Million
Telephones
4 Billion
XInternet
Electronic Chips
60 Billion
“XInternet” Beyond the PCForrester Research, May 2001
Revised 2007
CS 150  Spring 2007 – Lec #28 – Power  2
PC
Internet
X
Internet
Year
“XInternet” Beyond the PCCS 150  Spring 2007 – Lec #28 – Power  3
Forrester Research, May 2001
Ericsson T68
Cell PhonesCS 150  Spring 2007 – Lec #28 – Power  4
CS 150  Spring 2007 – Lec #28 – Power  5
CS 150  Spring 2007 – Lec #28 – Power  6
More bits/m3
Rapidly increasingtransistor density
Rapidly decliningsystem cost
Important (Wireless) Technology TrendsCS 150  Spring 2007 – Lec #28 – Power  7
CS 150  Spring 2007 – Lec #28 – Power  8
Rapid Growth: MachinetoMachine Devices (mostly sensors)
SpeedDistanceCostTradeoffs
Important (Wireless)Technology TrendsCS 150  Spring 2007 – Lec #28 – Power  9
CS 150  Spring 2007 – Lec #28 – Power  10
Power: Rate of work being done wrt time
Rate of energy being used
Watts = Joules/seconds
Units:
A very practical
formulation!
If we would like
to know total energy
CS 150  Spring 2007 – Lec #28 – Power  11
CS 150  Spring 2007 – Lec #28 – Power  12
1 Joule of Heat Energy per Second
20 W rating: Maximum power the package is able to transfer to the air. Exceed rating and resistor burns.
1 Watt
This is how electric tea pots work ...
Heats 1 gram of water 0.24 degree C
0.24 Calories per Second
+
1V

1 Ohm Resistor
CS 150  Spring 2007 – Lec #28 – Power  13
Like a resistor, iPod relies on passive transfer of heat from case to the air
Why? Users don’t want
fans in their pocket ...
To stay “cool to the touch” via passive cooling, power budget of 5 W
If iPod nano used 5W all the time, its battery would last 15 minutes ...
CS 150  Spring 2007 – Lec #28 – Power  14
14 hours for music,
4 hours for slide shows
Battery has 1.2 Whour rating:Can supply
1.2 W of power for 1 hour
85 mW for music
300 mW for slides
Powering an iPod nano (2005 edition)1.2 W / 5 W = 15 minutes
More Whours require bigger battery and thus bigger “form factor” 
it wouldn’t be “nano” anymore!
CS 150  Spring 2007 – Lec #28 – Power  15
20 hour battery life for audio,
6.5 hours for movies (80GB version)
Up from 14 hours for 2005 iPod nano
Up from 4
hours for 2005
iPod nano
24 hour
battery life for audio
5 hour battery life for photos
Thinner than 2005 iPod nano
12 hour
battery life
CS 150  Spring 2007 – Lec #28 – Power  17
Apple MacBook  Weighs 5.2 lbs
8.9 in
1 in
12.8 in
Performance: Must be “close enough” to desktop performance ... many people no longer own a desktop
Size and Weight: Ideal: paper notebook
Heat: No longer “laptops”  top may get “warm”, bottom “hot”. Quiet fans OK
CS 150  Spring 2007 – Lec #28 – Power  18
Almost full 1 inch depth. Width and height set by available space, weight.
Battery: Set by size and weight limits ...Battery rating: 55 Whour
At 2.3 GHz, Intel Core Duo CPU consumes 31 W running a heavy load  under 2 hours battery life! And, just for CPU!
46x energy than iPod nano. iPod lets you listen to music for 14 hours!
At 1 GHz, CPU consumes
13 Watts. “Energy saver” option uses this mode ...
CS 150  Spring 2007 – Lec #28 – Power  19
Toshiba Portege 3110 laptop  20%
Handspring PDA  10%
Nokia 61xx  33%
Battery TechnologyCS 150  Spring 2007 – Lec #28 – Power  20
55 Whour battery stores the energy of
1/2 a stick of dynamite.
If battery shortcircuits,
catastrophe is possible ...
CS 150  Spring 2007 – Lec #28 – Power  21
GPU
CPU
LCD
Backlight
LCD
CPU Only Part of Power Budget2004era notebook running a full workload.
If our CPU took no power at all to run, that would only double battery life!
CS 150  Spring 2007 – Lec #28 – Power  22
Machine rooms are expensive … removing heat dictates how many servers to put in a machine room.
Electric bill adds up! Powering the servers + powering the air conditioners is a big part of TCO
Reliability: running computers hot makes them fail more often
CS 150  Spring 2007 – Lec #28 – Power  23
Switch
Thermal Image of Typical Cluster RackM. K. Patterson, A. Pratt, P. Kumar, “From UPS to Silicon: an endtoend evaluation of datacenter efficiency”, Intel Corporation
CS 150  Spring 2007 – Lec #28 – Power  24
CS 150  Spring 2007 – Lec #28 – Power  25
CS 150  Spring 2007 – Lec #28 – Power  26
Switching Energy:
energy used to
switch a node
Calculate energy dissipated in pullup:
Energy supplied
Energy stored
Energy dissipated
An equal amount of energy is dissipated on pulldown
CS 150  Spring 2007 – Lec #28 – Power  27
number of nodes (or gates)
clock rate
Switching Power(probability of switching on
any particular clock period)
Therefore:
CS 150  Spring 2007 – Lec #28 – Power  28
Transistor drain regions
“leak” charge to substrate.
1020% of total chip power
~1nWatt/gate
few mWatts/chip
CS 150  Spring 2007 – Lec #28 – Power  29
This source of power consumption is becoming increasing significant as process technology scales down
For 90nm chips around 1020% of total power consumption Estimates put it at up to 50% for 65nm
Other Sources of Energy ConsumptionTransistor s/d conductance
never turns off all the way
Low voltage processes much worse
CS 150  Spring 2007 – Lec #28 – Power  30
Factors influencing power consumption:
n: total number of nodes in circuit
: activity factor (probability of each node switching)
f: clock frequency (does this effect energy consumption?)
Vdd: power supply voltage
What control do you have over each factor?
How does each effect the total Energy?
Controlling Energy Consumption: What Control Do You Have as a Designer?Our design projects do not optimize for power consumption
CS 150  Spring 2007 – Lec #28 – Power  31
Due to reduced V and C (length and width of Cs decrease, but plate distance gets smaller)
Recent slope reduced because V is scaled less aggressively
Scaling Switching Energy per GateMoore’s Lawat work …
CS 150  Spring 2007 – Lec #28 – Power  32
From: “Facing the Hot Chips Challenge Again”, Bill Holt, Intel, presented at Hot Chips 17, 2005.
(Pstandby) by raising Vt
We can increase speed
by raising Vdd andlowering Vt
We can reduce CV2 (Pactive)
by lowering Vdd
Device Engineers Trade Speed and PowerFrom: Silicon Device Scaling to the Sub10nm Regime
Meikei Ieong,1* Bruce Doris,2 Jakub Kedzierski,1 Ken Rim,1 Min Yang1
CS 150  Spring 2007 – Lec #28 – Power  33
From: “Facing the Hot Chips Challenge Again”, Bill Holt, Intel, presented at Hot Chips 17, 2005.
CS 150  Spring 2007 – Lec #28 – Power  34
Find enough tricks, and you can afford to raise Vdd a little so that you can raise the clock
speed!
Clock speed
unchanged ...
Design tricks: architecture & circuits
Lower Vdd, lower C,
but more leakage
Intel: Comparing 2 CPU Generations ...CS 150  Spring 2007 – Lec #28 – Power  35