Title: Power Estimation of Digital Systems
1Power Estimation of Digital Systems
22 September 2003
2Introduction
- Why power of nano-electronics became so
important? - Because of Moores law still holds true through
complex applications - Mobile systems battery bottleneck
- High performance computation heat extraction
- Operating cost and reliability
- Data warehouse of ISP with 8000 servers needs 2
MW - Power or Energy? Arent they go hand-in-hand?
- Power varies significantly with time!
- A given battery has fixed amount of energy
- Average power consumption Energy/Execution-time
- Decides average chip and junction temperature
- Decides battery life (if peak current lt rated
current) - Peak power and current
- Voltage drops, hot spots, rate of battery
discharge - Power-efficient, Energy-efficient,
Battery-efficient design paradigms do exist!
3Components of Power Consumption
- System hardware platform software (sys.
app.) - Software impacts hardware power consumption
- Static power
- Sub-threshold leakage reverse biased junction
leakage - Quiescent biasing power (in case of non-CMOS
circuits) - Dynamic power
- Charging and discharging of capacitance
(switching activity) - Short circuit power during transition (rate of
change, delay) - Alternative grouping (used at component/cell
level) - Switching power at the boundaries of cells
- Internal cell power
- Short circuit power
- Switching power at internal nodes
4System Abstractions - Power
Functional Specifications and Constraints System
Level Netlist Register Transfer Level (RTL)
Netlist Component/Cell Level Netlist Layout or
Configuration-bits Chip
Time complexity
Accuracy of power characterization
Opportunities for optimization
5Power Characterization
- Measurement (Chip/Board Level)
- Most accurate
- Perhaps the fastest, if setup and tools exist
- Too late to change hardware details
- Software/Load control is still possible
- Typically used for software optimizations
- Transistor Level (estimation)
- Spice simulation of transistor level netlist
- Most accurate in the simulation world
- Requires complete implementation details
- Unmanageable time complexity even for simpler
designs - Typically used for cell/component
characterization - Synopsys PowerMill (said to provide spice-like
accuracy)
6Power Characterization (cont)
- Cell Level (estimation)
- After logic synthesis
- Requires RTL implementation
- Simulation to capture switching activity
- Requires delay simulation if glitches need to be
accounted - Characterized cells empirical formulas or table
look-up - Interconnect power
- Either unaccounted or
- Using estimated wire load models (typically based
on experience) or - Extracted layout (if done after physical
synthesis) - Still unmanageable time complexity especially to
use in design space exploration - Synopsys PrimePower
- Netlist, interconnect capacitance, VCD traces,
cell power library
7Power Characterization (cont)
- Register Transfer Level (estimation)
- Requires conceptual RTL description (detailed
micro-architecture) - Data-path is modeled as netlist of macro cells,
which are characterized offline - Control path and glue logic
- Either unaccounted or estimated based on I/O
- Simulation to capture switching activity
- Typically glitches are not considered but methods
do exist - Interconnect power
- Typically unaccounted but possible to estimate
through floor-planning - Typically used in DSE mostly using in-house tools
- Synopsys PowerCompiler but too naive
- Synthesizable RTL, switching activity, power
library
8System Level Power Estimation
- For Design Space Exploration
- Least accurate but uncertainty of exploration
results can be reduced if models have good
fidelity - Purpose, target architecture and available system
details govern the system-level estimation models - Selecting algorithm or designing hardware for
given algorithm? - ASIC based or processor based?
- Is ISA fixed or extensible?
- Typically system-level power estimation models
are macro-architecture template specific - Major constituents of power consumption
- Computation, communication, storage units
peripherals
9Power Estimation Models
- Instruction Level Power Estimation
- First introduced to characterize processor power
consumption to drive software optimizations - Each instruction is associated with some current
- Inter instruction effects for better accuracy
- Later experiments on StrongARM by Amit Sinha
APC - Current/instruction 0.2A (averaged over all
instructions) - Min-max variation of 38 of average current
- Address mode and data dependent variation is
smaller - But, max current variation across benchmarks is lt
8 ! - Concluded that first order energy model of a
given processor is, E V I(V, f) T - Second order effects can be significant for
data-path dominated processors such as DSP, VLIW
10Power Estimation Models (cont)
- Second order effects are best characterized by
events which can be obtained from simulation - I ( W1 C1 W2 C2 ) I0(V, f)
- C denotes event frequency
- W denotes associated weight (obtained by
characterization) - Events can be either fine-grained or coarse
grained - Events can be classes of instructions or
- Micro-architecture component/sub-component
accesses (WATTCH) or - Events such as normal accesses, stalls,
- Power State Machines by Luca Benini et. al.
- States of the FSM represent mode of operation
- Active, Idle, Power off,
- Transitions are labeled with triggering events,
transition energy cost and perhaps transition
times - Abstract specification to generate events can be
accompanied - Useful to analyze power management schemes
11Power Estimation Models (cont)
- Integration of both the above models results in a
more powerful model - States of the FSM represent mode of operation
- Certain events cause state change and certain
dont - Power/energy consumption in each state can be
event and parameter dependent (empirical
formulas, table look-ups) - Read and write access of memory in active mode
- ADD and MUL instructions of a processor
- of transitions of a bus
- Transitions are labeled with triggering events,
transition energy cost and perhaps transition
times - Abstract specification to generate events can be
accompanied - Useful for both system-level and RTL power
estimation of wide variety of components - Trade-off of accuracy Vs characterization effort
is possible
12Questions?