Keynote Address I -- Some Real Observations on Virtual Machines -- Session 1A: Cache and Memory -- Replica Victim Caching to Improve Reliability of In-Cache Replication -- Efficient Victim Mechanism on Sector Cache Organization -- Cache Behavior Analysis of a Compiler-Assisted Cache Replacement Policy -- Modeling the Cache Behavior of Codes with Arbitrary Data-Dependent Conditional Structures -- Session 1B: Reconfigurable and Embedded Architectures -- A Configurable System-on-Chip Architecture for Embedded Devices -- An Auto-adaptative Reconfigurable Architecture for the Control -- Enhancing the Memory Performance of Embedded Systems with the Flexible Sequential and Random Access Memory -- Heuristic Algorithm for Reducing Mapping Sets of Hardware-Software Partitioning in Reconfigurable System -- Session 2A: Processor Architecture and Design I -- Architecture Design of a High-Performance 32-Bit Fixed-Point DSP -- TengYue-1: A High Performance Embedded SoC -- A Fault-Tolerant Single-Chip Multiprocessor -- Session 2B: Power and Energy Management -- Initial Experiences with Dreamy Memory and the RAMpage Memory Hierarchy -- dDVS: An Efficient Dynamic Voltage Scaling Algorithm Based on the Differential of CPU Utilization -- High Performance Microprocessor Design Methods Exploiting Information Locality and Data Redundancy for Lower Area Cost and Power Consumption -- Session 3A: Processor Architecture and Design II -- Dynamic Reallocation of Functional Units in Superscalar Processors -- Multiple-Dimension Scalable Adaptive Stream Architecture -- Impact of Register-Cache Bandwidth Variation on Processor Performance -- Session 3B: Compiler and Operating System Issues -- Exploiting Free Execution Slots on EPIC Processors for Efficient and Accurate Runtime Profiling -- Continuous Adaptive Object-Code Re-optimization Framework -- Initial Evaluation of a User-Level Device Driver Framework -- Keynote Address II -- A Generation Ahead of Microprocessor: Where Software Can Drive uArchitecture To? -- Session 4A: Application-Specific Systems -- A Cost-Effective Supersampling for Full Scene AntiAliasing -- A Simple Architectural Enhancement for Fast and Flexible Elliptic Curve Cryptography over Binary Finite Fields GF(2 m ) -- Scalable Design Framework for JPEG2000 System Architecture -- Real-Time Three Dimensional Vision -- Session 4B: Interconnection Networks -- A Router Architecture for QoS Capable Clusters -- Optimal Scheduling Algorithms in WDM Optical Interconnects with Limited Range Wavelength Conversion Capability -- Comparative Evaluation of Adaptive and Deterministic Routing in the OTIS-Hypercube -- A Two-Level On-Chip Bus System Based on Multiplexers -- Keynote Address III -- Make Computers Cheaper and Simpler -- Session 5A: Prediction Techniques -- A Low Power Branch Predictor to Selectively Access the BTB -- Static Techniques to Improve Power Efficiency of Branch Predictors -- Choice Predictor for Free -- Performance Impact of Different Data Value Predictors -- Session 5B: Parallel Architecture and Programming -- Heterogeneous Networks of Workstations -- Finding High Performance Solution in Reconfigurable Mesh-Connected VLSI Arrays -- Order Independent Transparency for Image Composition Parallel Rendering Machines -- An Authorization Architecture Oriented to Engineering and Scientific Computation in Grid Environments -- Session 6A: Microarchitecture Design and Evaluations -- Validating Word-Oriented Processors for Bit and Multi-word Operations -- Dynamic Fetch Engine for Simultaneous Multithreaded Processors -- A Novel Rename Register Architecture and Performance Analysis -- Session 6B: Memory and I/O Systems -- A New Hierarchy Cache Scheme Using RAM and Pagefile -- An Object-Oriented Data Storage System on Network-Attached Object Devices -- A Scalable and Adaptive Directory Scheme for Hardware Distributed Shared Memory -- Session 7A: Potpourri -- A Compiler-Assisted On-Chip Assigned-Signature Control Flow Checking -- A Floating Point Divider Performing IEEE Rounding and Quotient Conversion in Parallel -- Efficient Buffer Allocation for Asynchronous Linear Pipelines by Design Space Localization.