Chalmers University of Technology Power Estimation Flex Soc

  • Slides: 56
Download presentation
Chalmers University of Technology Power Estimation Flex. Soc Seminar Series – 2004 -03 -15

Chalmers University of Technology Power Estimation Flex. Soc Seminar Series – 2004 -03 -15 Daniel Eckerbert daniel. eckerbert@ieee. org Flex. So. C Seminar Series – 2004 -03 -15 Page 1

Chalmers University of Technology Outline • Why power estimation? • Power macromodeling • Future

Chalmers University of Technology Outline • Why power estimation? • Power macromodeling • Future directions for power estimation Most pictures (non-Matlab-plots) are courtesy of Intel Corp. Flex. So. C Seminar Series – 2004 -03 -15 Page 2

Chalmers University of Technology Why Power Estimation? (Conference -Presentation Answer) • Heat removal is

Chalmers University of Technology Why Power Estimation? (Conference -Presentation Answer) • Heat removal is expensive (fans, heat-sinks) • Energy stored in battery is limited • Power delivery is expensive (area, reliability, verification, packaging) Flex. So. C Seminar Series – 2004 -03 -15 Page 3

Chalmers University of Technology Why Power Estimation? (Flex. So. C Answer) • Compiler optimizations

Chalmers University of Technology Why Power Estimation? (Flex. So. C Answer) • Compiler optimizations • “Designing” range of chips for certain applications • ? ? ? You tell me! Flex. So. C Seminar Series – 2004 -03 -15 Page 4

Chalmers University of Technology Power Reduction Techniques • Activity reduction • Supply voltage scaling

Chalmers University of Technology Power Reduction Techniques • Activity reduction • Supply voltage scaling • Leakage reduction (cut-off techniques, stacking etc) But, by how much does the power of a specific design needs to be reduced? And which power mechnism constitutes a problem? Flex. So. C Seminar Series – 2004 -03 -15 Page 5

Chalmers University of Technology Power Dissipation Basics Flex. So. C Seminar Series – 2004

Chalmers University of Technology Power Dissipation Basics Flex. So. C Seminar Series – 2004 -03 -15 Page 6

Chalmers University of Technology From Where Does the Power Increase Stem? Flex. So. C

Chalmers University of Technology From Where Does the Power Increase Stem? Flex. So. C Seminar Series – 2004 -03 -15 Page 7

Chalmers University of Technology Increased Integration Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Increased Integration Flex. So. C Seminar Series – 2004 -03 -15 Page 8

Chalmers University of Technology Increased Integration 120 billion transistors per wafer!!! SRAM chips fabricated

Chalmers University of Technology Increased Integration 120 billion transistors per wafer!!! SRAM chips fabricated on a 300 mm wafer Flex. So. C Seminar Series – 2004 -03 -15 Page 9

Chalmers University of Technology Increased Density Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Increased Density Flex. So. C Seminar Series – 2004 -03 -15 Page 10

Chalmers University of Technology Power Density Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Power Density Flex. So. C Seminar Series – 2004 -03 -15 Page 11

Chalmers University of Technology Heat Removal (die) Flex. So. C Seminar Series – 2004

Chalmers University of Technology Heat Removal (die) Flex. So. C Seminar Series – 2004 -03 -15 Page 12

Chalmers University of Technology Heat Removal (package) Flex. So. C Seminar Series – 2004

Chalmers University of Technology Heat Removal (package) Flex. So. C Seminar Series – 2004 -03 -15 Page 13

Chalmers University of Technology Power Macro Modeling Flex. So. C Seminar Series – 2004

Chalmers University of Technology Power Macro Modeling Flex. So. C Seminar Series – 2004 -03 -15 Page 14

Chalmers University of Technology Architecture Level Power Macromodeling à la Wattch • P=0. 1*Psw(αmax)+Psw(α,

Chalmers University of Technology Architecture Level Power Macromodeling à la Wattch • P=0. 1*Psw(αmax)+Psw(α, state) Flex. So. C Seminar Series – 2004 -03 -15 Page 15

Chalmers University of Technology What Do We Want from a Power Estimation Methodology? •

Chalmers University of Technology What Do We Want from a Power Estimation Methodology? • Accurate • Fast • Provide information for power reduction Flex. So. C Seminar Series – 2004 -03 -15 Page 16

Chalmers University of Technology Levels of Power Estimation Flex. So. C Seminar Series –

Chalmers University of Technology Levels of Power Estimation Flex. So. C Seminar Series – 2004 -03 -15 Page 17

Chalmers University of Technology Why Use Macro Models? • Circuit simulations excessively time- and

Chalmers University of Technology Why Use Macro Models? • Circuit simulations excessively time- and memory-consuming • Designers need to run long traces to compare solutions (only possible using macro models) Flex. So. C Seminar Series – 2004 -03 -15 Page 18

Chalmers University of Technology Estimation Tool Run-Times Run-time HSpice* Power. Mill * Macro** 16

Chalmers University of Technology Estimation Tool Run-Times Run-time HSpice* Power. Mill * Macro** 16 b Han. Carlson 12 d 14 h 8 h 49 m (30 x) 6 m 43 s (2700 x / 80 x) 8 b Multiplier 29 d 6 h 17 h 37 m (40 x) 9 m 42 s (4300 x / 100 x) 32 b Multiplier N/A 14 d 10 h 2 h 2 m (N/A / 170 x) * Highly optimized code by team of software designers ** Highly unoptimized C++ code by one overworked circuit designer Flex. So. C Seminar Series – 2004 -03 -15 Page 19

Chalmers University of Technology Precision • Circuit simulations give full or close to full

Chalmers University of Technology Precision • Circuit simulations give full or close to full precision (depending on extraction) • Macro modeling can give range of precision levels (with a maximum precision determined by methodology) • Macro model precision is limited by the characterization Flex. So. C Seminar Series – 2004 -03 -15 Page 20

Chalmers University of Technology Power Estimation Flow • Characterization – Requires lower level simulations

Chalmers University of Technology Power Estimation Flow • Characterization – Requires lower level simulations – Has to support maximum precision – One-time only, can afford to be slow • Estimation – Macro-model only – Can support multiple levels of precision – Frequently run, has to be fast Flex. So. C Seminar Series – 2004 -03 -15 Page 21

Chalmers University of Technology Characterization Flex. So. C Seminar Series – 2004 -03 -15

Chalmers University of Technology Characterization Flex. So. C Seminar Series – 2004 -03 -15 Page 22

Chalmers University of Technology Estimation Flex. So. C Seminar Series – 2004 -03 -15

Chalmers University of Technology Estimation Flex. So. C Seminar Series – 2004 -03 -15 Page 23

Chalmers University of Technology Switching Power (Circuit Level) Flex. So. C Seminar Series –

Chalmers University of Technology Switching Power (Circuit Level) Flex. So. C Seminar Series – 2004 -03 -15 Page 24

Chalmers University of Technology Switching Power (Early Power Macro Models) Flex. So. C Seminar

Chalmers University of Technology Switching Power (Early Power Macro Models) Flex. So. C Seminar Series – 2004 -03 -15 Page 25

Chalmers University of Technology Switching Power (VLSI Research Group Style) Equation-based 0→ 1 tracking,

Chalmers University of Technology Switching Power (VLSI Research Group Style) Equation-based 0→ 1 tracking, even for intermediate nodes depending on accuracy • Physical model based on nodal capacitances and voltage swings • Enables semi-automatic characterization Flex. So. C Seminar Series – 2004 -03 -15 Page 26

Chalmers University of Technology Short-Circuit Power (Circuit Level) Flex. So. C Seminar Series –

Chalmers University of Technology Short-Circuit Power (Circuit Level) Flex. So. C Seminar Series – 2004 -03 -15 Page 27

Chalmers University of Technology Short-Circuit Power (Macro Model) Flex. So. C Seminar Series –

Chalmers University of Technology Short-Circuit Power (Macro Model) Flex. So. C Seminar Series – 2004 -03 -15 Page 28

Chalmers University of Technology Interconnect Modeling Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Interconnect Modeling Flex. So. C Seminar Series – 2004 -03 -15 Page 29

Chalmers University of Technology CRC Flex. So. C Seminar Series – 2004 -03 -15

Chalmers University of Technology CRC Flex. So. C Seminar Series – 2004 -03 -15 Page 30

Chalmers University of Technology RLC Flex. So. C Seminar Series – 2004 -03 -15

Chalmers University of Technology RLC Flex. So. C Seminar Series – 2004 -03 -15 Page 31

Chalmers University of Technology Subthreshold-Leakage Power Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Subthreshold-Leakage Power Flex. So. C Seminar Series – 2004 -03 -15 Page 32

Chalmers University of Technology Subthreshold-Leakage Power (Circuit Level) Flex. So. C Seminar Series –

Chalmers University of Technology Subthreshold-Leakage Power (Circuit Level) Flex. So. C Seminar Series – 2004 -03 -15 Page 33

Chalmers University of Technology Subthreshold-Leakage Power • Stacking effects • Long settling times •

Chalmers University of Technology Subthreshold-Leakage Power • Stacking effects • Long settling times • …? Flex. So. C Seminar Series – 2004 -03 -15 Page 34

Chalmers University of Technology Subthreshold-Leakage Power (Macro Model) • Equation-based model considering on- and

Chalmers University of Technology Subthreshold-Leakage Power (Macro Model) • Equation-based model considering on- and off-states of the transistors constituting the gate • Enables semi-automatic characterization • Possible extensions for stacking • Possible extensions for multiple clock-cycle settling times Flex. So. C Seminar Series – 2004 -03 -15 Page 35

Chalmers University of Technology Gate-Leakage Power Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Gate-Leakage Power Flex. So. C Seminar Series – 2004 -03 -15 Page 36

Chalmers University of Technology Oxide Thickness 12 Å Flex. So. C Seminar Series –

Chalmers University of Technology Oxide Thickness 12 Å Flex. So. C Seminar Series – 2004 -03 -15 Page 37

Chalmers University of Technology Oxide Thickness Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Oxide Thickness Flex. So. C Seminar Series – 2004 -03 -15 Page 38

Chalmers University of Technology Gate-Leakage Power (Circuit Level) • Enough equations and theory to

Chalmers University of Technology Gate-Leakage Power (Circuit Level) • Enough equations and theory to use up the entire Flex. So. C seminar series Flex. So. C Seminar Series – 2004 -03 -15 Page 39

Chalmers University of Technology Gate-Leakage Power (Macro Model) • Equation-based model considering on- and

Chalmers University of Technology Gate-Leakage Power (Macro Model) • Equation-based model considering on- and off-states of the transistors constituting the gate • Enables semi-automatic characterization • Complications include leakage paths originating in one gate and ending up in another gate Flex. So. C Seminar Series – 2004 -03 -15 Page 40

Chalmers University of Technology Separating Power Dissipation Mechanisms Flex. So. C Seminar Series –

Chalmers University of Technology Separating Power Dissipation Mechanisms Flex. So. C Seminar Series – 2004 -03 -15 Page 41

Chalmers University of Technology Leakage Power Increase Flex. So. C Seminar Series – 2004

Chalmers University of Technology Leakage Power Increase Flex. So. C Seminar Series – 2004 -03 -15 Page 42

Chalmers University of Technology Active vs. Leakage Power Flex. So. C Seminar Series –

Chalmers University of Technology Active vs. Leakage Power Flex. So. C Seminar Series – 2004 -03 -15 Page 43

Chalmers University of Technology Separation of Mechanisms (Rise- and Fall-Times) Flex. So. C Seminar

Chalmers University of Technology Separation of Mechanisms (Rise- and Fall-Times) Flex. So. C Seminar Series – 2004 -03 -15 Page 44

Chalmers University of Technology Separation of Mechanisms (Supply-Voltage Scaling) Flex. So. C Seminar Series

Chalmers University of Technology Separation of Mechanisms (Supply-Voltage Scaling) Flex. So. C Seminar Series – 2004 -03 -15 Page 45

Chalmers University of Technology Mismatch for Methodologies without a Leakage Component Flex. So. C

Chalmers University of Technology Mismatch for Methodologies without a Leakage Component Flex. So. C Seminar Series – 2004 -03 -15 Page 46

Chalmers University of Technology Mismatch for Methodologies without Leakage or Dynamica Frequency and Supply

Chalmers University of Technology Mismatch for Methodologies without Leakage or Dynamica Frequency and Supply Scaling Flex. So. C Seminar Series – 2004 -03 -15 Page 47

Chalmers University of Technology Added Complications Flex. So. C Seminar Series – 2004 -03

Chalmers University of Technology Added Complications Flex. So. C Seminar Series – 2004 -03 -15 Page 48

Chalmers University of Technology Complex Gates • Most circuit-level models are only valid for

Chalmers University of Technology Complex Gates • Most circuit-level models are only valid for a single transistor or an inverter • Macromodels have to be able to account for large components, 32 b multipliers etc • Complexity increases super-linearly with the number of transistors in the component Flex. So. C Seminar Series – 2004 -03 -15 Page 49

Chalmers University of Technology Non-Conforming Components There are classes of components which do not

Chalmers University of Technology Non-Conforming Components There are classes of components which do not conform to the presented basic models: • Clock generators • Memories • etc Flex. So. C Seminar Series – 2004 -03 -15 Page 50

Chalmers University of Technology Clock Generation (DLL) Flex. So. C Seminar Series – 2004

Chalmers University of Technology Clock Generation (DLL) Flex. So. C Seminar Series – 2004 -03 -15 Page 51

Chalmers University of Technology Clock Generators (Delay Element) Flex. So. C Seminar Series –

Chalmers University of Technology Clock Generators (Delay Element) Flex. So. C Seminar Series – 2004 -03 -15 Page 52

Chalmers University of Technology Non-Linear Frequency Dependence for Certain Types of Components Flex. So.

Chalmers University of Technology Non-Linear Frequency Dependence for Certain Types of Components Flex. So. C Seminar Series – 2004 -03 -15 Page 53

Chalmers University of Technology Dynamic Precision • Different architectures might impose different precision requirements

Chalmers University of Technology Dynamic Precision • Different architectures might impose different precision requirements for different components (static during estimation) • Different operating modes might warrant different need for precision (changes during estimation) Flex. So. C Seminar Series – 2004 -03 -15 Page 54

Chalmers University of Technology Conclusions • Power estimation is not simply about averaging the

Chalmers University of Technology Conclusions • Power estimation is not simply about averaging the current through the supplies • Circuit simulation is too slow • A lot of research is needed to enable highlevel power estimation for future designs: some in mechanism modeling but more importantly in the estimation framework Flex. So. C Seminar Series – 2004 -03 -15 Page 55

Chalmers University of Technology Power estimation is not as easy as it looks Flex.

Chalmers University of Technology Power estimation is not as easy as it looks Flex. So. C Seminar Series – 2004 -03 -15 Page 56