High-Level Power Analysis and Optimization presents a comprehensive description of power analysis and optimization techniques at the higher (architecture and behavior) levels of the design hierarchy, which are often the levels that yield the most power savings. This book describes power estimation and optimization techniques for use during high-level (behavioral synthesis), as well as for designs expressed at the register-transfer or architecture level.
High-Level Power Analysis and Optimization surveys the state-of-the-art research on the following topics: power estimation/macromodeling techniques for architecture-level designs, high-level power management techniques, and high-level synthesis optimizations for low power.
High-Level Power Analysis and Optimization will be very useful reading for students, researchers, designers, design methodology developers, and EDA tool developers who are interested in low-power VLSI design or high-level design methodologies.
Auflage
Softcover reprint of the original 1st ed. 1998
Sprache
Verlagsort
Zielgruppe
Für Beruf und Forschung
Research
Illustrationen
Maße
Höhe: 254 mm
Breite: 178 mm
Dicke: 12 mm
Gewicht
ISBN-13
978-1-4613-7481-7 (9781461374817)
DOI
10.1007/978-1-4615-5433-2
Schweitzer Klassifikation
1. Introduction.- 1.1 Low power design.- 1.2 Design abstraction and levels of the design hierarchy.- 1.3 Benefits of high-level power analysis and optimization.- 1.4 Book overview.- 2. Background.- 2.1 Sources of power consumption.- 2.2 Methods for reducing power and energy consumption.- 2.3 High-level design techniques.- 2.4 High-level synthesis application domains.- 3. Architecture-Level Power Estimation.- 3.1 Analytical power models.- 3.2 Characterization based activity and power macromodels.- 3.3 Power and switching activity estimation techniques for control logic.- 3.4 Conclusions.- 4. Power Management.- 4.1 Clock-based power management: Gated and multiple clocks.- 4.2 Pre-computation.- 4.3 Scheduling to enable power management.- 4.4 Operand isolation.- 4.5 Power management through constrained register sharing.- 4.6 Controller-based power management.- 4.7 Conclusions.- 5. High-Level Synthesis For Low Power.- 5.1 Behavioral transformations.- 5.2 Module selection.- 5.3 Resource sharing.- 5.4 Scheduling.- 5.5 Supply voltage vs. switched capacitance trade-offs.- 5.6 Optimizing memory power consumption during high-level synthesis.- 5.7 Reducing glitching power consumption during high-level design.- 5.8 Conclusions.- 6. Conclusions And Future Work.- References.
